Effect of Finishing and Polishing Direction On The Marginal Adapt

University of Iowa
Iowa Research Online

Theses and Dissertations
2011
Effect of finishing and polishing direction on the

marginal adaptation of resin-based composite
restorations in vitro
Laurie St-Pierre
University of Iowa
Copyright 2011 Laurie St-Pierre

This dissertation is available at Iowa Research Online: http://ir.uiowa.edu/etd/2774
Recommended Citation
St-Pierre, Laurie. "Effect of finishing and polishing direction on the marginal adaptation of resin-based composite restorations in vitro."
Master's thesis, University of Iowa, 2011.
http://ir.uiowa.edu/etd/2774.
Follow this and additional works at: http://ir.uiowa.edu/etd

Part of the Other Dentistry Commons
EFFECT OF FINISHING AND POLISHING DIRECTION ON THE

MARGINAL ADAPTATION OF RESIN-BASED COMPOSITE
RESTORATIONS IN VITRO
by
Laurie St-Pierre
A thesis submitted in partial fulfillment

of the requirements for the
Master of Science degree in Operative Dentistry
in the Graduate College of
The University of Iowa
December 2011
Thesis Supervisors: Associate Professor Cathia Bergeron

Professor Marcos A. Vargas
Copyright by
LAURIE ST-PIERRE
2011
All Rights Reserved
Graduate College
The University of Iowa
Iowa City, Iowa
CERTIFICATE OF APPROVAL
______________________________
MASTERS THESIS
___________________
This is to certify that the Masters thesis of

Laurie St-Pierre
has been approved by the Examining Committee
for the thesis requirement for the Master of Science degree in
Operative Dentistry at the December 2011 graduation.
Thesis Committee:
____________________________________________
Cathia Bergeron, Thesis Supervisor
_____________________________________________
Marcos A. Vargas, Thesis Supervisor
_____________________________________________
Deborah S. Cobb
_____________________________________________
Ma. Marcela Hernndez
_____________________________________________
Justine L. Kolker
_____________________________________________
Fang Qian
To my fiance, ric, for his unconditional support and love despite the
distance
To my parents for layering the foundation of determination and for their
encouragements
To my mentors, for their help, support and willingness to teach anytime
ii
ACKNOWLEDGMENTS
I would like to express my sincere appreciation to my mentors, Dr.

Cathia Bergeron and Dr. Marcos A. Vargas, for their constant availability,
precious guidance, time and effort put throughout this project.
I would also like to thank all my committee members: Dr. Deborah S.
Cobb, Dr. Ma. Marcela Hernndez, Dr. Justine L. Kolker and Dr. Fang Qian
for their availability, invaluable assistance and suggestions.
Special thanks to the Central Microscopy Research Facility at the
University of Iowa and its staff: Kathy Walters, Jean Ross and Chantal
Allamargot to have taught me and allowed me to use the field emission
scanning electron microscope for this research project.
I thank Maggie Hogan from the Dows Institute for Dental Research for
her help in teeth collection and for her suggestions.
I would like to thank Laval University as well as my colleagues Dr.
Sylvie Morin, Dr. Sylvie Breton and Dr. Denis Robert for believing in me and
giving me the opportunity to fulfill my professional goal.
Finally, I thank the faculty, the staff and my precious classmates in
the Department of Operative Dentistry and the Department of Family
Dentistry at the University of Iowa. You are all truly wonderful. Special
thanks to Dr. Deborah S. Cobb, graduate program director, and Dr. Gerald
Denehy who made me feel at home and part of the Iowa Family.
iii
TABLE OF CONTENTS
LIST OF TABLES ............................................................................................. vii

LIST OF FIGURES .......................................................................................... viii
LIST OF ABBREVIATIONS ............................................................................... x
CHAPTER I - INTRODUCTION ........................................................................ 1
Purpose of the Study .......................................................................... 2
Research Hypotheses ......................................................................... 3
Main Hypotheses ......................................................................... 3
Secondary Hypotheses ................................................................. 4
CHAPTER II - LITERATURE REVIEW ............................................................ 5
Enamel and Dentin Composition ...................................................... 5
Adhesion to Enamel and Dentin ....................................................... 8
Acid-Etching Technique .............................................................. 8
Dental Adhesives ....................................................................... 10
Resin-Based Composites .................................................................. 13
History and Evolution ............................................................... 13
Formulation and Chemistry ...................................................... 14
Organic Polymer Matrix ..................................................... 14
Inorganic Filler Particles .................................................... 15
Coupling Agent .................................................................... 16
Initiator-Accelerator:........................................................... 17
Classification of Resin-Based Composites ................................ 17
Reasons for Failure .......................................................................... 18
Dental Plaque ................................................................................... 20
Definition and Composition....................................................... 20
Accumulation of Plaque on Resin-Based Composite
Restorations ............................................................................... 21
Effect of Bacterial Leakage around Restorations on the
Pulp ............................................................................................ 22
Preparation ....................................................................................... 23
Finish Lines ............................................................................... 23
Marginal Adaptation of Resin-Based Composites .......................... 26
Polymerization Shrinkage ......................................................... 27
Hygroscopic Expansion .............................................................. 29
Finishing and Polishing Techniques ............................................... 29
Effect of Finishing and Polishing Procedures on Marginal
Adaptation ........................................................................................ 31
Delayed vs. Immediate Polishing .............................................. 32
Effect of Heat ............................................................................. 34
Wet vs Dry Finishing .......................................................... 35
Effect of the Type of Resin-Based Composite on Finishing and
Polishing ........................................................................................... 36
Polishability and Polish Retention ........................................... 36
iv
Physical and Mechanical Properties ......................................... 40

Effect of the Type of Resin-Based Composite on the
Marginal Adaptation ................................................................. 41
Direction of Finishing ...................................................................... 42
Measurement of Marginal Seal ....................................................... 45
Microleakage .............................................................................. 45
Marginal Adaptation ................................................................. 46
Scanning Electron Microscopy ......................................................... 47
Evaluation of Marginal Adaptation Using SEM ............................ 48
Quantitative Margin Analysis .................................................. 48
Summary .......................................................................................... 49
CHAPTER III - MATERIALS AND METHODS.............................................. 51
General Design ................................................................................. 51
Specimens Preparation .................................................................... 51
Teeth Sectioning ........................................................................ 52
Enamel Preparation .................................................................. 53
Standardized Preparation ......................................................... 54
Specimens Restoration ..................................................................... 58
Groups ........................................................................................ 59
Adhesion to Enamel ................................................................... 60
Restoration ....................................................................................... 62
Restoration Placement .............................................................. 62
Finishing and Polishing ............................................................ 62
Sof-Lex ................................................................................. 63
Diamond Bur, Sof-Lex disc and Rubber Polishers ............. 63
Verification of Polishing and Margins ...................................... 65
Specimens Preparation for FeSEM Observation ............................ 66
Fabrication of Replicas .............................................................. 66
Mounting of Replicas, Sputter Coating .................................... 68
FeSEM Observation .................................................................. 68
Images Preparation .......................................................................... 70
Evaluation of Marginal Quality ...................................................... 70
Intra-Observer Reliability ............................................................... 72
Statistical Methods .......................................................................... 77
Variables .................................................................................... 77
Power Analysis .......................................................................... 77
Statistical Analysis .................................................................... 77
Summary .......................................................................................... 79
CHAPTER IV - RESULTS ................................................................................ 84
Intra-Observer Reliability Measurements ...................................... 84
Descriptive Findings for Marginal Quality Criteria....................... 84
Difference between Two Polishing Directions ................................ 87
Continuous Margins (MQ1) .................................................... 87
Marginal Irregularities (MQ2) ............................................... 88
Gaps (MQ3+MQ4) ................................................................... 88
Difference between Two Polishing Techniques............................... 90
Gaps (MQ3+MQ4) ................................................................... 91
Difference between Two Types of Resin-Based Composite ............ 93

Gaps (MQ3+MQ4) ................................................................... 94
Difference between the Eight Experimental Groups...................... 96
Differences in Continuous Margins (MQ1) ............................ 96
Differences in Marginal Irregularities (MQ2) ....................... 97
Differences in Gaps (MQ3+MQ4) ........................................... 98
Interaction between Resin-Based Composite and Polishing
Technique for Each Polishing Direction.......................................... 99
1. Interaction Within Polishing Direction C-T ..................... 100
2. Interaction Within Polishing Direction T-C ..................... 100
Interaction between Polishing Technique and Polishing
Direction for Each Resin-Based Composite .................................. 101
1- Interaction Within FSU .................................................... 102
2- Interaction Within PT4 ..................................................... 103
Interaction between Resin-Based Composite and Polishing
Direction for Each Polishing Technique ....................................... 104
1- Interaction Within Polishing Technique SL .................... 104
2- Interaction Within Polishing Technique R ...................... 106
Interaction between Resin-Based Composites, Polishing
Technique and Polishing Direction ............................................... 107
For Variable Continuous Margin (MQ1) .............................. 108
For Variable Marginal Irregularity (MQ2) .......................... 109
For Variable Gaps (MQ3+MQ4) ........................................... 110
Statements on Research Hypotheses ............................................ 112
CHAPTER V - DISCUSSION ......................................................................... 113
Marginal Adaptation vs. Microleakage ......................................... 114
Quantitative Margin Analysis ....................................................... 115
Replicas Technique ........................................................................ 117
Control of Psychomotor Skills Variability..................................... 118
Variables Affecting Gap Formation............................................... 118
Cavity Preparation and Restoration ............................................. 119
Effect of Finishing and Polishing Direction .................................. 120
Effect of Finishing and Polishing Technique ................................ 122
Effect of Type of Resin-based Composite ...................................... 124
Differences between the Eight Experimental Groups .................. 125
Interactions .................................................................................... 126
Control of Possible Bias ................................................................. 128
Strengths of the Study ................................................................... 129
Limitations of the Study ................................................................ 130
Clinical Significance....................................................................... 131
Suggestions for Future Research .................................................. 133
CHAPTER VI - CONCLUSION ...................................................................... 135
REFERENCES ................................................................................................ 137
vi
LIST OF TABLES
Table 1. Marginal Quality Criteria .................................................................. 72

Table 2. Resin-Based Composite Specifications ............................................... 81
Table 3. Adhesive Specifications ...................................................................... 82
Table 4. Sof-Lex Finishing and Polishing Discs Specifications ....................... 82
Table 5. HiLuster Rubber Polishers Specifications ......................................... 83
Table 6. Descriptive Statistics of Marginal Quality by Treatment
Groups .............................................................................................. 86
Table 7. Comparison in Marginal Quality Difference between Two
Polishing Directions under Different Conditions ........................... 89
Table 8. Comparison in Marginal Quality Difference between Two
Polishing Techniques under Different Conditions.......................... 92
Table 9. Comparison in Marginal Quality between Two Resin-based
Composites under Different Conditions .......................................... 95
Table 10. Mean MQ1 by Experimental Groups ............................................... 97
Table 11. Mean MQ2 by Experimental Groups ............................................... 98
Table 12. Mean MQ3+MQ4 by Experimental Groups ..................................... 99
Table 13. Summary of the Full Model Used in the Random Effects in
Mixed Model ANOVA for MQ1 ...................................................... 109
Mixed Model ANOVA for MQ2 ...................................................... 110
Mixed Model ANOVA for MQ3+MQ4 ............................................ 111
Table 16. Statements on Main Hypotheses.................................................... 112
Table 17. Statements on Secondary Hypotheses ........................................... 112
vii
LIST OF FIGURES
Figure 1. Field emission scanning electron microscope image of

prepared enamel etched with 37% phosphoric acid for 20
seconds. The result is a partial dissolution of enamel prisms
and interprismatic substance. ......................................................... 10
Figure 2. Classification of dental adhesives..................................................... 11
Figure 3. A 90cavosurface angle is parallel to enamel rods whereas a
45 bevel and a 30 bevel expose the end of enamel rods and
allow a smooth transition in shade and texture between the
resin-based composite and the tooth structure. .............................. 25
Figures 4. Particle size and shape are very different between Filtek
Supreme Ultra and Point 4. ............................................................. 40
Figure 5. Human molar sectioned along its mesio-distal axis using a
precision saw machine (Isomet 1000, Buehler, Lake Bluff, IL,
USA).................................................................................................. 52
Figure 6. Sections of a molar before preparation and restoration. ................. 53
Figure 7. Enamel was flattened using silicon carbide grinding paper on
a polishing machine (Rotopol-V, Struers, Cleveland OH,
USA).................................................................................................. 54
Figure 8. Flattened enamel without dentin exposure. .................................... 54
Figure 9. Standardization of preparation in order to obtain two 30
bevels on each specimen. Having set a width of 3 mm,
calculations were made to determine the exact depth of
0.87 mm. ........................................................................................... 55
Figure 10. Bevel preparation using a flat-end fine diamond bur
(8847KR.31.016 Brasseler, USA) with water coolant..................... 56
Figures 11. Preparations width and depth measured with a digital
caliper (Model 47256, Cen-Tech, Harbor Freight Tools,
Camarilla, CA, USA). ....................................................................... 56
Figures 12. FeSEM micrographs of samples prepared without enamel
flattening. The outline of the preparation is curved and
undefined. ......................................................................................... 57
Figures 13. FeSEM micrographs of samples prepared with enamel
flattening. The outline of the preparation is straight and well
defined. ............................................................................................. 58
viii
Figure 14. Group assignment and number of specimen according to the

type of resin-based composite, the finishing and polishing
procedure and the direction. ............................................................ 59
Figure 15. On each specimen, both margins were finished and polished
with the same technique, one from the resin-based composite
to the tooth structure (C-T) (left margin) and the other from
the tooth structure to the resin-based composite (T-C) (right
margin). ............................................................................................ 60
Figures 16. Adhesion procedure using OptiBond FL....................................... 61
Figure 17. Placement of resin-based composite in one increment with
careful adaptation to minimize excess using an IPC
instrument (Brasseler, USA) and a brush (2 ROYAL SoftGrip SG3010, China) ........................................................................ 63
Figure 18. Sof-Lex group. Finishing and polishing done with a series of
Sof-Lex XT discs (3M ESPE, St-Paul, MN, USA). .......................... 64
Figures 19. Second polishing technique. .......................................................... 64
Figure 20. Light microscope (Zeiss, Thornwood, NY, USA) used to verify
the quality of the polishing surface ................................................. 65
Figure 21. Specimens preparation for FeSEM observations. .......................... 67
Figures 22. Sputter coating of samples prior to observation under
FeSEM. ............................................................................................. 69
Figures 23. Initial observation of replicas at low magnification to
localize the margins. These two images also show the 2 mm
outline marks done with a caliper prior to sputter coating and
the vertical central mark intended to separate and localize
the two margins................................................................................ 69
Figure 24. Marginal adaptation measured with ImageJ software
(ImageJ 1.44p, Wayne Rasband, National Institute of Health,
USA). The length of the entire studied margin was measured
as well as the length of any artifact and all defects belonging
to each marginal quality criteria. .................................................... 71
Figures 25. Marginal quality 1 criterion (MQ1). ............................................. 73
Figure 29. Mean values with the standard deviation for continuous
margin (MQ1), marginal irregularities (MQ2) and gaps
(MQ3+MQ4) among the eight experimental groups expressed
in % of entire margin length studied. .............................................. 85
ix
LIST OF ABBREVIATIONS
FSU:
Filtek Supreme Ultra (3M ESPE, St-Paul, MN, USA)
PT4:
Point 4 (Kerr, Orange, CA, USA)
SL :
Polishing technique using a series of Sof-Lex discs
R:
Polishing technique using a fine diamond finishing bur followed

by a dark orange Sof-Lex disc and rubber polishing cups.
C-T:
Finishing/polishing direction from resin-based composite resin to

tooth structure
T-C:
Finishing/polishing direction from tooth structure to resin-based

composite
FeSEM: Field emission scanning electron microscope

SEM:
Scanning electron microscope
1
CHAPTER I
INTRODUCTION
Within the last decades, dental patients expectations toward esthetic

restorations have greatly increased. Different tooth-colored materials such as
ceramic and resin-based composites have become very popular to restore
teeth when esthetic results are needed. Currently, resin-based composite is
one of the most widely used materials in esthetic dentistry. A great
advantage of this material is its ability to strongly bond to enamel and
dentin. It can be applied with minimal thickness and be used to create
ultraconservative restorations maintaining sound tooth structure. Thus, it
helps preserving pulp vitality and tooth resistance to fracture.
Polishing of resin-based composites is an important step to improve
esthetics and longevity of these types of restorations (Yap et al., 1998;
Jefferies, 1998; Baratieri L. N. et al., 1998; Powers et al., 2006). Several
consequences may result from a lack of polishing. First, porosities and
scratches as well as poor marginal adaptation can lead to staining and
discoloration of the restoration which may compromise esthetics and
longevity. Inadequate polishing can also create uneven surfaces and facilitate
dental plaque retention which may increase risks of gingivitis and recurrent
caries. For all those reasons, polishing procedures help to maintain longevity
of restorations and preserve good oral health.
2
However, studies (Yu et al., 1990; Brackett et al., 1997; Yap et al.,
1998; Lopes et al., 2002; Schmidlin and Gohring, 2004; Yalcin et al., 2006;
Cenci et al., 2008; Maresca et al., 2010) have shown that finishing and
polishing procedures themselves may negatively affect the marginal
adaptation of resin-based composite and create a marginal deficiency
between the tooth structure and the restorative material. Clinically, this
defect can be seen as a white margin immediately after polishing and may
lead to marginal staining over time (Fukushima et al., 1988; Schmidlin and
Gohring, 2004). Other possible problems may also occur, such as dental
plaque and debris accumulation into the defect which will increase risks of
recurrent
caries.
Moreover,
marginal
deficiency
can
create
communication with the underlying dentin and trigger tubular fluid

movement, resulting in post-operative sensitivity and ultimately pulpal
irritation.
Although resin-based composite finishing and polishing procedures are
well documented, there is little information in the dental literature regarding
the orientation of polishing procedures and its impact on marginal integrity.
Purpose of the Study

The purpose of this study was to assess the effect of polishing direction
on the marginal adaptation of resin-based composite restorations when
enamel is beveled. Resin-based composite restorations were polished in two
different directions: from the resin-based composite to the tooth structure
3
and from the tooth structure to the resin-based composite. The outcome was
assessed on two resin-based composite materials: a nanofilled (Filtek
Supreme Ultra, 3M ESPE) and a microhybrid (Point 4, Kerr, Orange, CA,
USA). This study also evaluated the effect of the type of polishing system.
Comparisons were made between Sof-Lex XT discs (3M ESPE) and rubber
polishing cups (Hi Luster, Kerr, Bioggio, Switzerland) for the two directions
mentioned above. The findings of this study may lead to important guidelines
on the best way to polish resin-based composite restorations in order to
prevent marginal defects that can be detrimental to restorations longevity.
Research Hypotheses
Main Hypotheses
Null Hypothesis (1): there is no difference in the marginal adaptation

between polishing direction C-T and polishing direction T-C under
different conditions.
Null Hypothesis (2): there is no difference in marginal adaptation

between polishing technique SL and polishing technique R under
different conditions.
Null Hypothesis (3): there is no difference in marginal adaptation

between resin-based composite FSU and resin-based composite PT4
under different conditions.
4
Secondary Hypotheses
o Null hypothesis (1.2): Resin-based composite has no influence on how
polishing technique affects marginal adaptation within each polishing
direction.
o Null hypothesis (2.2): Polishing technique has no influence on how
polishing direction affects marginal adaptation within each resin-based
composite.
o Null hypothesis (3.2): Resin-based composite has no influence on how
polishing direction affects marginal adaptation within each polishing
technique.
o Null hypothesis (4.2): there is no interaction among the three factors:
resin-based composites, polishing techniques and polishing directions.
5
CHAPTER II
LITERATURE REVIEW
Enamel and Dentin Composition

Enamel is the hardest and most mineralized tissue found in the
human body (Ten Cate, 1985). Mature enamel is often called a calcium
deficient carbonate containing apatite (Stoodley et al., 2008). It is mainly
composed of an inorganic structure (95-98% by weight) in which the primary
constituent is hydroxyapatite (Summitt, 2006). Because of its high mineral
content, enamel is very brittle and can be demineralized in acids leading to
dental caries or erosion (Ten Cate, 1985). The principle of solubility in acids
is exploited in dental adhesion (Powers et al., 2006). Enamel is not vital and
not sensitive. Therefore it cannot be regenerated after structure loss (Ten
Cate, 1985). Enamel also contains, in a low proportion, an organic content (12% by weight) and water (4% by weight) (Summitt, 2006). Enamel is
organized in enamel rods or prisms which are oriented perpendicular to the
external surface of the tooth and parallel between them (Summitt, 2006). The
rods are separated by an interprismatic substance (Ten Cate, 1985). After
tooth eruption, enamel undergoes maturation which increases its resistance
to demineralization (Lopes et al., 2007). During this maturation process,
minerals present in saliva fill interprism spaces (Lopes et al., 2007). Being
semi-permeable, enamel is able to exchange ions with its environment (Ten
Cate, 1985). Therefore, fluoride ions present topically in fluoride water,
6
toothpaste or other professional treatment enter the enamel structure to
form fluorohydroxyapatite and fluoroapatite which lower its solubility in
acids (Ten Cate, 1985; Stoodley et al., 2008). The external surface of the
enamel is often hypermineralized and presents high fluoride content
(Kanemura et al., 1999; Albers, 2002).
Dental enamel contains a feature known as striae of Retzius which run
perpendicular from dentino-enamel junction to the outer surface of enamel.
They are the result of enamel formation and represent growth lines. On the
surface, these striae of Retzius form a parallel pattern of lines and
depressions called perikymata (Ten Cate, 1985).
It has been shown that enamel may contain an aprismatic outer layer.
In 1966, Ripa et al. (1966) showed that the outer 30 m layer of 70% of
permanent teeth and 100% of deciduous teeth is composed of aprismatic
enamel which differs from the underlying enamel where crystals are oriented
perpendicular to the enamel surface. The authors evaluated 60 erupted and
40 unerupted permanent teeth in which an aprismatic layer covered the
gingival third in 57% of cases, the middle third in 36% of cases and the entire
surface in 7% of cases.
Another study conducted by Whittaker (1982) confirmed the presence
of aprismatic enamel on the outer surface in both deciduous and permanent
teeth with a depth of less than 80m in most of the teeth evaluated.
Aprismatic enamel was more frequent in deciduous teeth than permanent
7
teeth. It was also noted that the surface area of aprismatic enamel gradually
increases from incisor to molars in both permanent and deciduous dentition
and changes between different areas of a same tooth. With time, the outer
enamel surface undergoes changes and the aprismatic layer is lost in areas of
attrition and wear (Whittaker, 1982; Ten Cate, 1985).
Except for its aprismatic form, enamel is homogeneous and maintains
a parallel orientation of enamel rods (Summitt, 2006).
Dentin differs from enamel in its organic content (18% by weight),
which is rich in type I collagen and in its water content (12% by weight). The
inorganic fraction containing hydroxyapatite is 70% by weight. Dentin is
composed of several dentinal tubules going through its entire thickness from
the pulp to the dentino-enamel junction. These tubules contain fluids and
odontoblasts, which are considered a prolongation of the pulp tissue.
Compared to enamel, dentin is heterogeneous and composed of peritubular
and intertubular dentin. The peritubular dentin is hypermineralized whereas
the intertubular dentin contains more collagen fibers. The width of these
tubules varies through the depth of the dentin. Tubules are more dense
(45 000/mm2) and wider close to the pulp (2.8 m) and less dense
(20 000/mm2) and narrower close to the dentino-enamel junction (0.8 m)
(Summitt, 2006).
8
Adhesion to Enamel and Dentin
Acid-Etching Technique
In 1955, Dr. Michael Buonocore (1955) made a remarkable advance in
dentistry with his discovery of the acid-etching technique. Using phosphoric
acid in vitro, he realized that the bond of an acrylic restoration to enamel was
significantly increased by about 100 fold. Based on these promising results,
Dr. Buonocore tried this method on upper and lower central incisors of
volunteer patients. A drop of acrylic resin was applied on treated and
untreated enamel surfaces. For untreated surfaces, the bond lasted for an
average of 11 hours as opposed to treated surfaces where the bond lasted for
more than 1070 hours. When the drop of acrylic placed over the acid-treated
enamel surface was mechanically removed, the enamel surface appeared
cleaned, opaque and white but recovered completely in few days. He
attributed the enhancement in bond longevity to an increase in surface area
and wettability, which allows a more intimate contact between the acrylic
and the tooth structure. Those findings were an important step forward in
conservative dentistry and in caries prevention because it led to the
possibility of sealing pits and fissures on occlusal surfaces (Buonocore, 1955;
Bergenholtz et al., 1982).
In his study, Buonocore (1955) used 85% phosphoric acid for 30
seconds. In the early clinical use, the recommended etching time was 60
seconds. Barkmeier et al. (1986) then showed that reducing etching time
from 60 seconds to 15 seconds with 37% phosphoric acid did not reduce shear
9
bond strength and etched enamel displayed a similar microporosity pattern
for both 15 and 60 seconds. Currently, the etching time suggested is at least
15 seconds (Summitt, 2006) and this is what most manufacturers recommend
(Lopes et al., 2007). However, on intact enamel where an aprismatic layer
might be present, longer etching time is recommended (Marshall et al.,
1975).
Acid-etching of enamel usually removes about 10 m of enamel and
leaves an irregular surface with microporosities of 5-50 m deep (Gwinnett,
1971; Summitt, 2006; Lopes et al., 2007). Acid-etching will also increase the
surface energy and wettability, which will facilitate the infiltration of a resin
material into the microporosities. After polymerization, this infiltrated resin
forms micromechanical interlocking or retentive resin tags, which are
responsible for a strong, resistant and long-lasting bond (Buonocore et al.,
1968; Summitt, 2006; Lopes et al., 2007).
The most reliable bond is achieved when the ends of enamel rods are
etched, whereas a less retentive bond is obtained when the sides of enamel
rods are etched (Albers, 2002). This concept depends on the tooth preparation
and will be discussed in the section named finish lines.
Observed under a field emission scanning electron microscope, the
pattern of etched enamel can appear as keyholes because of the preferential
dissolution of the interprismatic substance (Summitt, 2006) (Figure 1).
10
Figure 1. Field emission scanning electron microscope image of prepared

enamel etched with 37% phosphoric acid for 20 seconds. The result is a
partial dissolution of enamel prisms and interprismatic substance.
Dental Adhesives
Different classifications of dental adhesives exist in the literature
(Summitt, 2006). One classification is described in terms of adhesive
generations and refers to the sequence of their introduction on the market.
Another classification is based on the adhesion mechanism and the number
of steps for the bonding procedure. This classification describes total-etch
adhesives, self-etch adhesives and glass ionomers. For total-etch and selfetch adhesives, the classification is further subdivided simplified and nonsimplified applications as described by the number of steps (Figure 2). Totaletch products remove the outer layer of enamel and the smear layer produced
while instrumenting the tooth structure, whereas self-etching adhesives
dissolve those components (Summitt, 2006). Total-etch three-step adhesives
11
(non-simplified) are still currently considered as the gold standard (Swift,
2010).
Total-etch
3 steps
2 steps
Self-etch
2 steps
1 step
Self-etching
Primer
Primer
All-in-one
Primer + Adhesive
Self-etching
primer + adhesive
Adhesive
Adhesive
Figure 2. Classification of dental adhesives.

_______________________________________________________________________
Adapted from Summitt J.B., Robbins J.W., Hilton T.J., Schwartz R.S. (2006).
Fundamentals of Operative Dentistry: A Contemporary Approach, Illinois
Quintescence Books.
Because the outer enamel surface is hypermineralized, has high

fluoride content and may be aprismatic, it has been suggested to remove
12
0.1 mm of that enamel before acid-etching to provide better shear bond
strength compared to unprepared enamel (Schneider et al., 1981).
However, Kanemura et al. (1999) found no significant difference in
tensile bond strength between ground and intact enamel when using
phosphoric acid. Conversely, a self-etching primer on intact enamel resulted
in a poorly demineralized surface.
Adhesion to dentin is very different than adhesion to enamel mainly
because of the organic content of dentin and its variable degree of
permeability. Moreover, because the tubules become wider and more dense
close to the pulp and also because dentin is a vital tissue undergoing changes
overtime (affected dentin, sclerotic dentin, tertiary dentin) dentin adhesion is
more variable and less predictable than enamel adhesion. After tooth
preparation, a smear layer consisting of tooth debris covers the dentin
surface and protrudes in dentinal tubules. This smear layer is either
dissolved by a self-etching adhesive or completely removed with a total-etch
adhesive. To this day, the total-etch technique is still the most reliable, even
though the two-step self-etch technique is very promising. With total-etch
three-step adhesives, phosphoric acid is applied on dentin no longer than 15
seconds and rinsed thoroughly. This removes the smear layer and dissolves
the hydroxyapatite, leaving a network of collagen fibers and opened dentinal
tubules. This layer left slightly moist is further hybridized by the application
of an amphiphilic primer. The solvent contained in the primer is then
13
evaporated and a hydrophobic resin is applied and light-cured (Summitt,
2006).
Resin-Based Composites
History and Evolution
Introduced in 1878, silicates were the first tooth-colored dental
materials used for direct esthetic restorations followed by acrylic resin
materials in 1948 (Albers, 2002; Powers et al., 2006). Silicates are formed
through a reaction between alumina glass powder and phosphoric acid-based
liquid. Silicate and acrylic resin restorations were associated with early
failures because of their high coefficient of thermal expansion, low color
stability and weak adhesion. Silicates had additional limitations, such as
solubility and disintegration in the oral environment. Because of these
weaknesses, researchers sought a better material with greater physical
properties (Bowen, 1956).
Approximately six months after Buonocores discovery of the enamel
acid-etching technique, Dr. Rafael Bowen (1956) published an article
reporting the development of a new dental material, an epoxy resin, with
improved properties to overcome the weaknesses of silicates and acrylic
resins. This was the first step in discovering what is currently known as BisGMA, which was introduced in 1962 by Dr. Bowen. Dr. Bowen added quartz
fillers to that epoxy resin and, as a result, several properties were
significantly improved such as thermal expansion, adhesive strength, color
14
stability to ultraviolet light, insolubility in oral environment and material
stability (Bowen, 1956).
Resin-based composites are now known as reinforced polymers. The
first resin-based composite formulated was self-cured and resulted from the
mixture of a base and a catalyst. Because of the mixing process, voids and
porosities could be introduced in the material and compromise the esthetic
and the strength of the material. In 1970, resin-based composites photoactivated by ultraviolet (UV) wavelength were introduced. They were later
replaced by visible light-cured resin-based composites which are now widely
used as direct restorative materials to replace missing tooth structure.
(Powers et al., 2006)
Formulation and Chemistry

Resin-based composites are made of four components: an organic
polymer matrix, inorganic filler particles, a coupling agent and an initiatoraccelerator (Powers et al., 2006). Each these components influences the
mechanical and physical properties of the material (Asmussen and
Peutzfeldt, 1998).
Organic Polymer Matrix

The organic polymer matrix is composed of a diacrylate oligomer which
can be aromatic or urethane based. Oligomers found in most resin-based
composites are bisphenylglycidyl dimethacrylate (Bis-GMA) or urethane
15
dimethacrylate (UDMA). These oligomers present a high viscosity that needs
to be reduced for handling purposes and to allow addition of filler particles.
This is accomplished by adding diluent monomers, usually triethylene glycol
dimethacrylate (TEGDMA) (Powers et al., 2006; Floyd and Dickens, 2006;
Goncalves et al., 2009). In addition to decreasing the viscosity of Bis-GMA or
UDMA, TEGDMA also increases the depth of cure of resin-based composite.
However, because of its low molecular weight and its high conversion rate,
polymerization shrinkage is increased (Floyd and Dickens, 2006). In order to
reduce polymerization shrinkage, Bisphenol-A dimethacrylate (Bis-EMA), an
alternative diluent with high molecular weight, is present in some resinbased composite formulation replacing completely or partially TEGDMA
(Goncalves et al., 2009).
In an attempt to further reduce the polymerization shrinkage of resinbased composites, a new oligomer named silorane, has been developed and
introduced on the market (Powers et al., 2006; Goncalves et al., 2009).
Inorganic Filler Particles

The filler particles can come from different inorganic substances such
as quartz, glass, colloidal silica, zirconia, ytterbium trifluoride, aluminum
oxide, silicon dioxide, etc. Glass or quartz particles are usually of fine size,
whereas microfine particles can be obtained from colloidal silica. Depending
on the type of the particles and how they are processed, their shape can be
irregular or spherical. Filler particles features including size, shape, content
16
and distribution influence the physical and mechanical properties of resinbased composites (Powers et al., 2006; Summitt, 2006; Berger et al., 2011)
such as wear and polishability (Ergucu and Turkun, 2007; Marghalani,
2010). It has been shown that a smaller average particle size allows a better
polish and that irregularly shaped particles increase the surface roughness of
the material (Van Noort and Davis, 1984; Marghalani, 2010). A high filler
load is difficult to obtain with microfine particles without affecting the
handling properties. Some manufacturers have overcome this issue by first
loading a resin to a maximum capacity and, after polymerization, grinding
the material to obtain particles of various sizes. These prepolymerized
particles are then added as fillers in a resin matrix containing microfine
particles (Van Noort and Davis, 1984).
Coupling Agent
The coupling agent is a silane which role is to make a bond between
inorganic filler particles and the organic matrix. Filler particles are coated
with silane before mixing them to the unreacted organic matrix. The purpose
of the coupling agent is to transfer stresses to the particles in order to
improve strength and mechanical properties of resin-based composites.
(Powers et al., 2006)
17
Initiator-Accelerator:
The initiator in light-cured resin-based composites is a molecule
present in the organic matrix which initiates the polymerization reaction
after being activated by a light of 470 nm wavelength. Camphoroquinone is
the most commonly used initiator but because of its yellow shade, other
initiators are sometimes used for esthetic considerations. An accelerator, an
organic amine with a carbon double bond, is also incorporated in the organic
matrix. (Powers et al., 2006)
Classification of Resin-Based Composites

There are many different types of resin-based composites for a range of
different applications (Powers et al., 2006). Some types of resin-based
composites are more suitable for esthetic restorations whereas some other
types are made for occlusal stress bearing areas in posterior teeth.
In general, resin-based composites are categorized based on their
average particle size and the percentage of filler content (Powers et al.,
2006). The different types of resin-based composite are macrofilled,
microfilled, hybrids and nanofilled composites.
The
first
resin-based
composite
marketed
was
macrofilled.
Macrofilled resins have large particles of spherical shape ranging from 20 to

30 m in diameter. These materials are no longer used in dentistry.
Microfilled resins are composed of 0.04 to 0.2 m particles and contain pre-
18
polymerized particles of 10 to 20 microns. Because small particles have a
high surface area, they significantly increase the viscosity of the resin as
they are loaded in the organic matrix. When the material becomes too
viscous, its handling is inadequate for clinical use. The addition of prepolymerized particles in the microfilled resin matrix allows for a higher filler
content (up to 50-60%) while maintaining proper viscosity. Hybrid resinbased composites are made of a blend of irregular glass, quartz or zirconia
particles of microfine size (0.04 to 0.2 m) with fine particles (0.4 to 3 m). In
recent years, manufacturers have created particles in the submicron range
and therefore hybrids can now be called microhybrids or nanohybrids
depending on whether the particles are measured in microns or in
nanometers. Microhybrids have a filler content of 77 to 84% by weight.
Nanofilled resin-based composites are relatively new and consist of small
silica particles mixed with 0.6 m zirconia-silica nanoclusters (Powers et al.,
2006). Because the range of particle size is between 20nm and 70nm and the
average size is less than 100 nm, these materials are considered true
nanofilled as opposed to nanohybrids.
Reasons for Failure

The reasons for restoration failure are important to consider in order
to find the weaknesses of a material (Bernardo et al., 2007). Knowing the
weaknesses leads to advancements in dental materials. The adhesive
19
procedure when restoring with resin-based composite is very technique
sensitive and is affected by contamination (saliva, blood, plasma, water, etc.)
which may greatly affect the restorations longevity (Powers and Farah,
2010). A rubber dam isolation is required whenever possible to obtain a dry
operative field and avoid contamination by saliva, blood, crevicular fluid or
even patients breath, which can decrease the bond strength (Albers, 2002).
Besides the material itself, material handling, isolation field and patients
oral hygiene are among factors that may be responsible for restoration
failure (Bernardo et al., 2007).
In a survey among Finnish dentists, Forss et al. (2004) observed that
the most frequent reason for resin-based composites failure was recurrent
caries followed by tooth or restoration fracture and loss of the restoration.
Other reported causes of failure included marginal leakage, sensitivity and
inadequate esthetic.
Mjr et al. (1997) conducted a survey among general practitioner in
Denmark to assess the marginal failure of amalgam and resin-based
composite restorations. The survey revealed that recurrent caries was the
most common reason for marginal failure of resin-based composite
restorations. Non-carious marginal defects and marginal discolorations each
represented one fifth of all replacements of resin-based composite
restorations. However, in most cases of recurrent caries, the lesion was
located at the gingival margin.
20
In a randomized controlled clinical trial, Bernardo et al. (2007) also
found that recurrent caries was the main reason for resin-based composite
failure followed by restoration fracture. The restorations were placed in
permanent posterior teeth of teenagers and followed over seven years.
Dental Plaque
Definition and Composition
Dental plaque can be described as a soft and complex biological
material found on teeth surfaces. It is a biofilm composed mostly of organic
material such as food debris, mucin, dead epithelial cells and bacteria in a
matrix of polysaccharides, proteins, carbohydrates and lipids (Jablonski,
1982). The inorganic component contains calcium, phosphate and trace
amount of potassium, magnesium and sodium (Jablonski, 1982). Bacteria
present in dental plaque metabolize carbohydrates in alimentation to
produce acetyl-coa. In an anaerobic environment found in well established
dental plaque, glycolysis happens through fermentation and produces acids
(lactic, acetic, formic, propionic, pyruvic and succinic) that dissolve mineral
components of teeth and lead to dental caries (Simain et al., 2010). The
virulence of these cariogenic bacteria depends on their acid production
(Simain et al., 2010), their acid resistance and their capacity to strongly
adhere to teeth surfaces (Michalek et al., 1975; Emilson and Krasse, 1985;
Simain et al., 2010). Among these bacteria, Streptococcus mutans and
Lactobacilli have been shown to be associated with dental caries (Emilson
21
and Krasse, 1985; Simain et al., 2010). Other aerobic and anaerobic species
have been isolated from dental plaque, such as Corynebacterium, Neisseria,
Pseudomonas aeruginosa, fusobacteria, staphylococcus species, actinomyces

and other cocci (Simain et al., 2010). Recent studies using DNA hybridization
and polymerase chain reaction (PCR) have allowed the isolation of more
species and research is still active to find stronger association between
bacteria and dental caries (Russell, 2009).
Dental
plaque
is
additionally
associated
with
gingivitis
and
periodontal diseases. Moreover, due to its acidity, dental plaque triggers the
precipitation of calcium and phosphate ions in saliva into calculus found on
teeth surfaces.
Accumulation of Plaque on
Resin-Based Composite Restorations
It has been shown in a study conducted by Skjrland (1973) that
different restorative dental materials have various degrees of plaque
adhesion. This study combined an in vitro and an in situ evaluation of plaque
adhesion. In the in vitro portion of the study, the authors incubated discs of
dental materials (amalgam, silicates and resin-based composites) with S.
mutans in a nutrient broth (5% sucrose). The in situ portion of the study
consisted of evaluating appliances containing samples of dental materials
after exposure to a solution of 15% sucrose. The in vitro and in situ
experiments showed that plaque was formed in larger quantities on resin-
22
based composites compared to silicates and amalgams. Therefore, it was
concluded that dental plaque is more likely to adhere to resin-based
composites than to other dental restorative materials.
Zalkind et al. (1998) reported that finishing and polishing procedures
are important to reduce plaque accumulation on resin-based composites. The
purpose of that study was to compare finishing procedures (diamond burs,
tungsten burs and Sof-Lex discs) to resin-based composites (microfilled and
macrofilled) cured against a Mylar matrix on the adhesion of Streptococcus
mutans. The authors compared also the accumulation of Streptococcus

mutans on burnished vs. polished amalgam and on enamel samples. All the
resin-based composite specimens were cured against a Mylar matrix before
finishing
and
polishing
procedures.
Resin-based
composites
showed
statistically higher bacterial accumulation than amalgam samples but no

significant difference with enamel samples or resin-based composite cured
against a Mylar matrix. This demonstrates the importance of finishing and
polishing procedures. No significant difference was found among the
finishing procedures used for resin-based composite.
Effect of Bacterial Leakage around

Restorations on the Pulp
An in vivo study on monkeys assessed the effect of marginal leakage
on the dental pulp (Bergenholtz et al., 1982). The authors found that bacteria
from dental plaque can penetrate marginal gaps around restorations, reach
23
dentinal tubules and grow, potentially causing pulpal inflammation.
Histological analyses showed that bacterial penetration can result in
localized abscesses in the pulp tissue. The reaction of the pulp may include
the formation of reparative dentin or inflammation. The results of this study
suggested that marginal integrity around dental restorations is an important
factor to preserve pulpal health.
Brnnstrom (1984) also stated that bacteria may invade dental
tubules through a gap around restorations if they are open to the oral
environment causing early caries and pulpal inflammation.
Preparation
Resin-based composite offers the possibility of making ultraconservative restorations, preserving sound tooth structure. The outline of
the preparation should be limited to the defect that needs to be removed
(Summitt, 2006).
Finish Lines
Finish lines refer to the border of the preparation or the junction of the
prepared and unprepared surfaces of the tooth (Summitt, 2006). Different
types of finish lines exist (90 cavosurface angle, bevel, chamfer, shoulder,
etc.) and their use depends on the restorative material as well as the esthetic
needs.
24
A 90 cavosurface angle refers to an angle of 90 formed by the
junction of the walls of the preparation and the external surface of the tooth
(Sturdevant et al., 1985). It is also known as a butt joint and it is the most
desirable finish line for amalgam restorations in order to respect the material
requirements. Traditionally, it was believed that the ends of enamel rods
needed to be exposed before acid-etching procedures with bonded restorations
such as resin-based composites. Therefore a bevel was recommended for
retention and esthetics (Baratieri and Ritter, 2005). Early laboratory studies
comparing 90 cavosurface angle to beveling enamel were in favor of bevel in
order to reduce marginal leakage (Speiser and Kahn, 1977; Porte et al., 1984;
Crim et al., 1984). However, because of total acid-etching techniques and
overall improvements in adhesion and materials, it was recently suggested
that bevels may not be necessary for retention and prevention of
microleakage, but maybe just for esthetic reasons (Baratieri and Ritter,
2005). Microleakage is the infiltration of bacteria, oral fluids and other
materials such as debris at the tooth-restoration interface (Jablonski, 1982).
Bagheri et al. (2008) also reported that there was no difference in
microleakage between beveled and non-beveled preparations for resin-based
composite restorations. It has further been advised that bevels should not be
prepared in all situations since they eliminate sound tooth structure and also
because of the wide variety of resin-based composite shades available that
facilitate the creation of esthetic and conservative restorations (Baratieri and
Ritter, 2005).
25
Even though some evidence suggests that non-beveled preparations
enable a more conservative approach with similar esthetic results, bevels are
still commonly suggested (Baratieri and Ritter, 2005). Beveled finish lines
are more esthetic than 90 cavosurface margins due to the shade transition
between the resin-based composite and the tooth structure (Albers, 2002). A
45 bevel is a common finish line (Albers, 2002) but in esthetic zones, a
longer, irregular and scalloped bevel may be needed in order to create a
harmonious and esthetic blend between the resin-based composite and the
tooth structure (Summitt, 2006). Longer bevels can also be described by their
degree of inclination from the surface such as 30 bevels (Figure 3).
90 cavosurface angle
45 bevel
30 bevel
Figure 3. A 90cavosurface angle is parallel to enamel rods whereas a 45
bevel and a 30 bevel expose the end of enamel rods and allow a smooth
transition in shade and texture between the resin-based composite and the
tooth structure.
26
Marginal Adaptation of Resin-Based Composites
Longevity and esthetics of resin-based composite restorations depend
greatly on marginal integrity (Porte et al., 1984) which is the degree of
proximity between a restorative material and the tooth structure (Jablonski,
1982).
The minimal space between the restorative material and the tooth
structure allowing for bacterial infiltration has been reported to be about 0.51.0 m wide (Taylor and Lynch, 1992). However, even smaller gaps can cause
toxin and bacterial product infiltration, with harmful effect on tooth
structure (Taylor and Lynch, 1992).
The dental literature is quite controversial regarding the association
between gap sizes and recurrent caries. Some in vitro studies have
demonstrated a positive relationship between gap formation and recurrent
caries. However, when taking into account other factors present in an oral
environment, it seems that a gap does not influence the development of
recurrent caries.
Totiam et al. (2007) assessed the relationship between gap size and
recurrent caries. The authors observed different gap sizes ranging from 0 m
to 1,000 m (0 m, 25 m, 50 m, 250 m, 500 m, 1,000 m) and found that
wider gaps led to a more recurrent caries. Regardless of the significant
difference between gap sizes, all of the gaps studied showed development of
recurrent caries.
27
Conversely, in an in situ study, Cenci et al. (2008) evaluated the effect
of microleakage at the tooth-restoration interface when fluoride from
toothpaste or glass ionomer restorative material was present. The authors
did not show a positive association between gaps and recurrent caries.
Using a microcosm biofilm model, Cenci et al. (2009) evaluated the
relationship between different gap sizes (0 m, 50 m, 100 m, 180 m,
250 m) and recurrent caries using two restorative materials (resin-based
composite and glass ionomer). The authors concluded that the gap size
influences recurrent caries, but not in the presence of fluoride.
Patients oral hygiene has also been reported to be a primary factor
affecting recurrent caries formation while marginal quality of restorations
does not necessarily affect the longevity of restorations (Sarrett, 2007).
Many factors affect gap formation such as polymerization shrinkage,
hygroscopic expansion, heat and finishing and polishing procedures.
Polymerization Shrinkage
The polymerization process of resin-based composite results from a
cross-linking reaction that reduces spaces between monomer molecules. An
effect of this reaction is polymerization shrinkage which depends on the
oligomer and the diluents present in the material. When more diluents are
present, more shrinkage occurs. Microfilled resin-based composites, because
of their high resin matrix content, show 2-3 % shrinkage, whereas
28
microhybrid resin-based composites shrink on average from 0.6 to 1.4 %
(Powers et al., 2006).
Polymerization shrinkage can create stresses up to 13 MPa in the
material. This stress can be transferred to the tooth-restoration bonded
interface, disrupting marginal seal. This can lead to gap formation and
marginal leakage (Powers et al., 2006; Ferracane, 2008), post-operative
sensitivity, marginal staining or recurrent caries (Schneider et al., 2010).
Polymerization shrinkage can be affected by the cavity size and shape,
and by the flow of unset material during its polymerization (Taylor and
Lynch, 1993). The level of stress generated during polymerization shrinkage
can be estimated by the configuration factor, also known as C-factor,
described by Feilzer et al. (1987). The C-factor is the ratio of bonded to
unbonded surfaces. The higher the C-factor, the higher the stresses
developed in the material after its polymerization.
Different strategies have been proposed to reduce polymerization
shrinkage. Among these approaches, it has been suggested to control the
curing light irradiance at the beginning of the polymerization process while
the material is in a gel phase with soft-start technique, pulse-delay
method (Schneider et al., 2010) or step curing (Ferracane, 2008). Other
methods focus on the material manipulation during its insertion by
recommending a layering technique, the use of liners or by preheating the
resin-based composite (Schneider et al., 2010). However, a study using finite
element stress analysis reported that a layering technique may result in
29
higher residual stress than bulk insertion of resin-based composite (Versluis
et al., 1996). Some new resin-based composite materials are based on
silorane monomers instead of traditional methacrylates. Silorane monomers
are obtained by the compound of siloxane and oxiranes and their
polymerization is based on a ring-opening reaction that minimizes shrinkage
(Guggenberger and Weinmann, 2000).
Hygroscopic Expansion
Resin-based composites absorb water over time. This water sorption
phenomenon causes swelling of the material known as hygroscopic expansion
(Powers et al., 2006). Hygroscopic expansion may counteract the effect of
polymerization shrinkage, but at a slower pace (Feilzer et al., 1990; Taylor
and Lynch, 1993; Yap et al., 2003). Yap et al. (2003) reported that
hygroscopic expansion may improve the marginal seal of resin-based
composite by reducing marginal gaps associated with polymerization
shrinkage.
Finishing and Polishing Techniques

Finishing and polishing are important steps when restoring with
resin-based composites because they provide anatomical and physiological
contours to the restoration, hence increase tolerance of periodontal tissue and
resistance from staining and recurrent caries (Jefferies, 1998; Baratieri L. N.
et al., 1998). Finishing and polishing are essential to ensure proper esthetics
30
and good longevity of resin-based composite restorations (Jefferies, 1998).
Because a lack of finishing/polishing can lead to marginal staining,
discoloration of the restoration, gingival irritation and recurrent caries due to
plaque accumulation, finishing/polishing methods help achieving our primary
goal: to maintain good oral health (Yap et al., 1998; Jefferies, 1998; Baratieri
L. N. et al., 1998; Powers et al., 2006).
Finishing refers to the contouring, shaping and smoothing of the
restoration to give anatomical contours and to remove excess at the toothrestoration interface (Jefferies, 1998; Summitt, 2006; Jefferies, 2007).
Polishing is a step done after finishing when the surface gains a high luster
and an enamel-like texture (Summitt, 2006; Jefferies, 2007).
Finishing and polishing procedures of resin-based composites are
based on a controlled and selective wear of the material. Therefore, finishing
and polishing are done with different kinds of abrasives which are integrated
in different instruments. Aluminum oxide, also known as alumina, is often
used as well as carbide compounds, diamond, silicon dioxides, zirconium
oxide and zirconium silicate. These abrasives can be bonded to discs and
strips, impregnated in elastomeric material to make rubber cups, points and
wheels or incorporated into polishing pastes. Stones, diamond burs and
fluted carbide burs are also available for finishing. All these finishing and
polishing tools are offered in different degrees of abrasiveness. Brushes
impregnated with abrasives are fairly recent and can also be used for
polishing. (Jefferies, 2007)
31
These instruments are used in a sequence, working gradually toward
finer grits (Jefferies, 1998; Baratieri L. N. et al., 1998). For instance, 100 m
abrasive particles are needed to remove excess and to contour the
restoration. Scratches resulting from the contouring are then smoothed down
with abrasive particles of less than 100 m but higher than 15-20 m. The
final polishing is obtained with 20 m to 0.3 m particles (Jefferies, 1998).
Effect of Finishing and Polishing

Procedures on Marginal Adaptation
Finishing
and
polishing
procedures
may
affect
the
marginal
adaptation of resin-based composites (Yu et al., 1990; Brackett et al., 1997;

Yap et al., 1998; Lopes et al., 2002; Schmidlin and Gohring, 2004; Yalcin et
al., 2006; Cenci et al., 2008; Maresca et al., 2010). Defects at the toothrestoration interface can be seen as a white margin immediately after
polishing and may result in marginal staining over time (Fukushima et al.,
1988; Schmidlin and Gohring, 2004). A lack of adaptation of the resin-based
composite at the margin increases the risk for recurrent caries, postoperative sensitivity, pulpal irritation and staining of margins (Brannstrom,
1984; Yap et al., 1998).
Ritter (2005) has suggested that resin-based composite placement
should be done carefully before polymerizing the material to limit the need
32
for finishing. Minimal finishing helps maintaining the seal of the restoration
and preventing microcracks.
Delayed vs. Immediate Polishing

Polymerization of light-cured resin-based composite begins when the
material is exposed to a wavelength in a range of 450-500 nm (Summitt,
2006). Even if the material appears hard and fully cured after exposing it to
the curing light for the recommended time, the polymerization continues for
24 hours. Immediate finishing and polishing may cause plastic deformation
because only 75% of the material is cured after 10 minutes (Powers et al.,
2006). Therefore, it has been suggested to delay finishing and polishing
procedures until 24 hours after light curing the material (Lopes et al., 2002).
Asmussen et al. (1972) also proposed to delay any finishing procedures until
after hygroscopic expansion occurs because of the risk of fracture of the
unsupported enamel surrounding the gap. However, this older study only
considered
the
gap
formation
caused
by
polymerization
shrinkage,
disregarding the possible effect of finishing and polishing after hygroscopic

expansion occurred.
On the other hand, it has also been recommended to perform finishing
and polishing procedures immediately after curing. This statement was made
based on the fact that hygroscopic expansion will improve marginal
adaptation by closing the gap formed by polymerization shrinkage and
finishing/polishing procedures (Yap et al., 1998).
33
Lopes et al. (2002) reported that the benefits of delayed versus
immediate finishing were material and tooth structure dependent. This
study compared Sof-Lex discs used dry versus fine and extra-fine diamond
finishing burs used with water coolant and microfilled versus hybrid resinbased composite. In their study, the authors found no difference in enamel
margins. There was less leakage in dentin margins with delayed finishing
when a microfilled resin-based composite was used. The hybrid resin-based
composite did not show any difference for enamel or dentin margins as well
as for delayed versus immediate finishing. This study evaluated class V
restorations on human molars with a 90 cavosurface angle.
Cenci et al. (2008) assessed the effect of polishing methods, as well as
immediate versus delayed polishing, on surface roughness, microhardness
and microleakage of microfilled (Filtek A-110) and hybrid (Filtek Z-250)
resin-based composites after one year storage. This in vitro study used 256
freshly extracted bovine incisors where class V cavity preparations were
prepared with 90 cavosurface angle. The gingival margins were in dentin
and the occlusal margins in enamel. The authors tested three finishing and
polishing procedures: Sof-Lex XT discs, rubber-polishing cups followed by
Enamelize paste and sequential use of rubber-polishing cups, polishing
brushes and felt-polishing discs (Flexibuff) with Enamelize paste. Half of the
restorations were polished immediately after curing and half were delayed
for two weeks. The authors found no difference in microleakage among all
the polishing techniques evaluated. More leakage occurred in dentin margins
34
when the polishing was delayed. After one year storage, enamel margins
showed a decrease in sealing ability for both immediate and delayed
polishing. The authors of that study recommended immediate polishing of
resin-based composite.
This was supported by Yap et al. (1998) who assessed the effect of
immediate versus delayed finishing on marginal adaptation in enamel and
dentin. In this in vitro study, resin-based composite restorations using four
bonding systems were made on human molars with class V cavity
preparations. The preparations were in enamel and in dentin, with 90
cavosurface angles. Based on the results, the authors recommended that
finishing procedures should be done immediately after curing the material.
They also stated that delayed finishing is not advised and may be harmful to
marginal integrity. Delayed finishing procedures produce stress that may
cause a breakdown in the marginal seal obtained after hygroscopic
expansion. However, the effect of delayed finishing on marginal adaptation
depends on the tooth structure involved (enamel vs. dentin) and on the
adhesive system used.
Effect of Heat
Because resin-based composites are poor thermal conductors, heat
generated during polishing is confined to the outer layer and is not
transmitted into the bulk of the material. It has been suggested that heat
caused by dry polishing of resin-based composite can create a temperature
35
raise above glass transition temperature forming a hard surface layer which
increases mechanical properties of the material such as micro-hardness and
abrasion resistance (Davidson et al., 1981).
Nevertheless,
finishing
and
polishing
procedures
generate
considerable amount of heat which can be detrimental to the adhesive bond

at the tooth-restoration interface especially when the margins of the
restoration are located in dentin (Baratieri L. N. et al., 1998). Yal et al.
(1998) also reported that finishing resin-based composite produces heat that
can affect the leakage.
In order to control for the heat produced during finishing and
polishing, some authors have suggested to finish and polish with water
coolant.
Wet vs Dry Finishing

According to Powers et al. (2006), finishing of resin-based composite
should be done carefully with water coolant to prevent disruption of the
restoration-tooth interface. Interestingly, not only the bonded interface may
suffer from heat generated during finishing and polishing procedures, but
also the bond between the particles and the surrounding matrix. Indeed, Van
Noort et al. (1984) observed cracks at the interface between particles and the
surrounding matrix when Sof-Lex discs were used and explained this finding
by the heat created during the polishing process.
36
Lopes et al. (2002) also assessed the influence of dry versus wet
finishing methods on marginal integrity of resin-based composites. The
authors found no difference in enamel margins regardless of the finishing
method or the type of resin-based composite (microfilled or hybrid) used.
However, for dentin margins, multiple comparisons showed that delayed wet
finishing using fine and extra-fine diamond burs produced superior results
than the other groups.
Because of the difficulty to assess the texture and luster of the
restorations surface while finishing with water, some authors advocate
finishing with diamond burs dry with an air spray (Baratieri L. N. et al.,
1998). According to the same authors, water may be used with rubber points
to reduce heat whereas discs (such as Sof-Lex from 3M ESPE or Super-Snap
from Shofu) can be used dry or with air and water spray.
Effect of the Type of Resin-Based Composite on

Finishing and Polishing
Polishability and Polish Retention
Polishability of resin-based composites relies on the filler particle size
and morphology, the filler loading, the type of filler and on the polishing
method and instruments (Van Noort and Davis, 1984; Barghi and Lind, 2000;
Ergucu and Turkun, 2007; Marghalani, 2010). Smaller particles in resinbased composites are essential to obtain good polishing results (Venhoven et
al., 1996). Visually better polish can be achieved with particle size smaller
37
than 6 microns (Van Noort and Davis, 1984). Small particles can be
incorporated into the resin matrix in fewer quantities and therefore, because
of the high matrix content, these materials can be polished to a higher luster
(Strassler, 1990). Polishing resin-based composites containing coarse
particles leaves dull and rough surfaces and a shiny surface is never
completely obtained (Venhoven et al., 1996). It has been shown that
irregularly shaped particles also increase the surface roughness of the
material (Marghalani, 2010).
Gloss and polish retention are affected over time by the wear process.
Resin-based composite wear is a complex phenomenon and a consequence of
chemical and mechanical degradation. It is affected by several factors such as
particle size and morphology, filler loading, inter-particle space, hydrolysis
degradation of coupling agent bond, and depth of cure (Wu et al., 1984;
Bayne et al., 1992; Ferracane and Marker, 1992; Venhoven et al., 1996;
Turssi et al., 2005; Ardu et al., 2009; 3M ESPE, 2010). Indeed, some solvents
such as alcohol and acids contained in foods and drinks can accelerate the
wear of dental composite by weakening its outer surface layer which is
further abraded by mechanical stresses such as occlusal forces and teeth
brushing (Wu et al., 1984; Strassler, 1990; Ardu et al., 2009). Large particles
also have a negative effect on the polish retention of resin-based composites.
With the wear of the resin matrix, large particles protrude more from the
surface and, having a longer cantilever, the particles are pulled out earlier
from the surface leaving deep and wide craters where the appearance of the
38
restoration is dull (Venhoven et al., 1996; 3M ESPE, 2010). Turssi (2005)
found that particles morphology also influence the wear process where
irregular filler particles are more wear resistant than spherical particles. In
addition, inter-particle distance may play a role in polish retention. It has
been first suggested by Jrgensen (1978) and later by Bayne (1992) that an
inter-particle distance of 0.1 m or less would be protective against wear
because of the high wear resistance of particles compared to the surrounding
matrix. According to Bayne (1992), smaller particle size and higher filler
content decrease inter-particle distance which helps wear resistance.
In the present study, a nanofilled resin-based composite, Filtek
Supreme Ultra (3M ESPE, St-Paul, MN, USA), and a mycrohybrid resinbased composite, Point 4 (Kerr, Orange, CA, USA), were used.
Filtek Supreme Ultra (3M ESPE, St-Paul, MN, USA) contains 20 nm
silica particles, 4 to 11 nm zirconia particles, as well as loosely agglomerated
particles named nanoclusters made of silica and zirconia. The filler content is
approximately 72.5% by weight and 55.6% by volume (3M ESPE, 2009). The
non-agglomerated/non-aggregated filler particles and the particles in the
nanoclusters are spherical (Berger et al., 2009).
The nanoclusters are the result of a sintering process which softens
the particles and allows them to bond together, creating dense and compact
fillers. This allows higher filler loading which improves physical properties
and wear resistance. This technology has been modified from the first
generation of Filtek Supreme to create loosely agglomerated particles that
39
can be separated individually during the wear process. Nanoclusters, in
Filtek Supreme Plus, have been shown to wear with a rate and a pattern
similar to the surrounding filled matrix. In Filtek Supreme Ultra, the filler
technology has been modified by decreasing sintering, which allows a wider
range of cluster sizes without changing the properties. The wear rate and
pattern of Filtek Supreme Ultra is even more similar to the surrounding
filled matrix than it is with Filtek Supreme Plus (3M ESPE, 2010).
Point 4 (Kerr, Orange, CA, USA) contains filler particles made of
barium, aluminoborosilicate glass and silicon dioxide with an average
particle size of 0.4 microns. The filler loading is 76% by weight and 57 % by
volume (Watanabe et al., 2008; Kerr Dental Sybron Dental Specialties).
Because microhybrids contain glass particles, their shape is irregular
(LeSage, 2007).
Filtek Supreme Ultra and Point 4 are similar in their filler loading but
very different in their particle size, type and morphology (Figures 4a and 4b).
Because of the small particle size of Filtek Supreme Ultra, high polishability
and good retention of that polish are expected (Ergucu and Turkun, 2007).
On the other hand, Point 4 is a microhybrid containing large and small
particles with an average size of 0.4 m. Therefore, its immediate polish is
expected to be good, but the long term polish retention may be compromised
when larger particles are removed from the surface as a result of wear.
40
a)
b)
Figures 4. Particle size and shape are very different between Filtek Supreme
Ultra and Point 4.
a) Filtek Supreme Ultra, a nanofilled, contains nanoclusters, small and
spherical particles.
b) Point 4, a microhybrid, contains large and small particles of irregular
shape.
Physical and Mechanical Properties

It is expected that different types of resin-based composites will
influence the marginal adaptation of the restoration. Physical and
mechanical properties of resin-based composites depend on different factors
such as the matrix, the filler content and the coupling between the matrix
and the filler particles (Asmussen and Peutzfeldt, 1998). Filler loading,
41
particle size and particle morphology can affect polymerization shrinkage,
depth of cure, coefficient of thermal expansion, hardness, stiffness, wear
resistance, fracture toughness and resistance to abrasion (Braem et al., 1989;
Kim et al., 2002; Summitt, 2006). The amount of fillers that can be
incorporated into the matrix depends on the particle shape (Kim et al., 2002).
More spherical filler particles can be incorporated than irregular particles,
and prepolymerized filler particles can be incorporated in a lower quantity
(Kim et al., 2002). High filler content will likely improve the marginal
adaptation of the restoration because it decreases the polymerization
shrinkage and improves fracture toughness and resistance to abrasion.
However, the filler content must not exceed a certain limit otherwise the
viscosity of the material increases it becomes unsuitable for clinical
manipulation (Summitt, 2006).
Effect of the Type of Resin-Based Composite

on the Marginal Adaptation
Yalcin et al. (2006) evaluated the microleakage associated with
different
polishing
techniques
(Super-Snap
Rainbow
and
Astropol/Astrobrush) on a nanofilled, a nanohybrid and a microhybrid resinbased composite. Class V cavity preparations with a 90 cavosurface angle
were prepared on both buccal and lingual surfaces with occlusal margins in
enamel and gingival margins in dentin. Finishing and polishing were
performed after 24 hours. No significant differences in microleakage were
42
observed in enamel margins regardless of the material used and the
polishing procedure. However, microleakage in dentin margins was
dependent on the material and the polishing system used. In dentin, margins
showed the following leakage from the worst (most leakage) to the best (least
leakage): Grandio (nanohybrid) > Artemis (microhybrid) > Filtek Supreme
(nanofill). Therefore, within the limitations of that study, a nanofilled resin
showed less microleakage than a microhybrid resin in dentin margins.
Direction of Finishing
Only one study conducted by Maresca et al. (2010) evaluated the
direction (parallel or perpendicular) of finishing procedure in relation to the
gap formation. Because there are various finishing instruments available on
the market, the authors assessed the impact of different grits and types of
finishing burs as well as the orientation of finishing procedures (parallel or
perpendicular)
on
the
marginal
integrity
of
resin-based
composite
restorations in vitro. To assess the outcome, the authors used scanning

electron microscopy. This study published in 2010 by Christina Maresca et
al. was the Masters thesis of Dr. Maresca in the department of Operative
Dentistry in Chapel Hill, North Carolina. In their research design, the
authors used 75 bovine teeth prepared by flattening the facial surface with
abrasive paper (320, 400, 600, 1200-grit) under running water (1 minute for
each cycle). The high luster of enamel was then obtained with a polishing
paste containing 5-micron diamonds. Then, they prepared a class V-like
43
cavity (3 X 3 X 2 mm) on that flat facial surface. The dimensions of the
preparation were verified using a digital caliper. They stored the
preparations in a chloramine solution. Preparations were rinsed, gently dried
and restored with resin-based composite placed in one increment with the
material slightly overfilling the preparation. Then, they cured the resinbased composite for 40 seconds and placed the specimens in deionized water
for a week. They randomly divided their study samples into 15 groups (5
specimens per group) with a positive control (restoration mechanically
polished) and a negative control (regular-grit diamond). Finishing was
performed using different grit of diamond and carbide burs one week after
polymerization of the resin-based composite. They monitored the pressure
(0.5 N) applied during finishing with a custom-made device. The authors
finished the restorations in a constant mesio-distal direction. After finishing
all restoration, they prepared the specimens for observation with scanning
electron microscopy. They found that diamond burs resulted in better
marginal integrity than carbide burs and that the negative control group had
the worst marginal adaptation. They also found that the use of fine, extrafine and ultra-fine diamond burs one after the other gave the best results
followed by ultra-fine, extra-fine and fine diamond burs used individually.
Regarding the orientation of the finishing procedure, the authors did not find
any relationship with the size of the marginal gap. The authors stated that
the sample size was a limitation in the study and recommended further
studies. The strengths of this study are that all the preparations were made
44
by only one operator who was calibrated before the study. Moreover, they
used a device to monitor the pressure applied during the finishing procedure.
The sample preparation was well standardized. To flatten the surface, they
embedded the specimen where only the facial surface slightly protruded and
they used abrasive paper for a specified amount of time. However, the study
has also some weaknesses. A class V cavity preparation has a high C-factor.
Thus, placing the resin-based composite in one increment may have
disrupted the adhesive bond and such disruption may have been interpreted
as a defect in the marginal integrity caused by the finishing procedure. The
sample size is also another factor that may have influenced the results of the
study. Seventy-five bovine teeth were used and randomly divided into 15
groups. That means that each group contained only 5 specimens, which may
not have been adequate to detect differences with a sufficient power level.
Also, finishing procedures were delayed for a week which does not represent
the clinical reality. The authors made a class-V restoration with a 90
cavosurface angle whereas in clinical situations, the enamel margin is often
beveled. The perpendicular direction of the finishing procedure was
performed using the same mesio-distal orientation, but no distinction was
made between finishing from the resin-based composite to the tooth structure
or from the tooth structure to the resin-based composite. Despite these facts,
this study is valuable, relevant and contributes to the knowledge on finishing
procedures and longevity of resin-based composite restorations.
45
Except for Yu et al. (1990), the studies described previously evaluated
the effect of finishing and polishing methods on microleakage or marginal
adaptation when a 90 cavosurface angle is used as a finish line. Clinically,
for anterior restorations, it is highly recommended to bevel the enamel
margins in order to create imperceptible transition between the restoration
and the tooth structure and to increase the retention. The literature provides
little information regarding the effect of finishing procedures on marginal
adaptation of resin-based composite on beveled enamel. Therefore, designing
an in vitro study where the finish line is a bevel would be relevant and
representative of the clinical situation. Such a study design is not without
challenges since there is no description in the literature regarding a possible
standardization of a beveled preparation for an in vitro study and it is
difficult to clearly identify the margin of the beveled finish lines after
restoration.
Measurement of Marginal Seal

There are two methods that have been developed and used to assess
the seal and adaptation of restorations: microleakage and marginal
adaptation.
Microleakage
Microleakage is the infiltration of bacteria, oral fluids, ions and other
materials such as debris between the tooth and the restoration (Kidd, 1976;
46
Jablonski, 1982). Microleakage can be measured by dye penetration, air
pressure, bacterial infiltration, radioisotope, neutron activation analysis,
electrochemical studies, thermal and mechanical cycling and chemical
tracers (Kidd, 1976; Taylor and Lynch, 1992). Dye penetration is the most
common method used to assess microleakage in the dental literature and it is
generally evaluated qualitatively using specific criteria.
Marginal Adaptation
Marginal adaptation is the degree of proximity between the
restoration and the tooth structure (Jablonski, 1982; Taylor and Lynch,
1993). Marginal adaptation can be assessed in vivo and in vitro.
In vivo, marginal adaptation can be evaluated clinically with

qualitative criteria using a mirror and a probe or with stone casts and
photographs (Taylor and Lynch, 1993). Fukushima et al. (1988) assessed the
marginal adaptation of resin-based composite restorations clinically using an
explorer according to the USPHS criteria (Alfa, Bravo, Charlie and Delta).
However, the smallest marginal defect clinically detectable was evaluated to
have a width of 120 m and a depth of 40 m. Leingelder et al. (1982) has
also reported this limitation stating that ledges can be detected clinically if
they are wider than 100 m. In order to evaluate smaller marginal defects,
studies have used light microscopy and scanning electron microscopy (Taylor
and Lynch, 1993). This latter method is non-destructive and the samples can
47
be conserved and observed in vitro and in vivo at different points in time and
after various manipulations or storage (Blunck and Zaslansky, 2011).
Scanning Electron Microscopy

For many years, scanning electron microscopy (SEM) has been used to
describe dental structures and pathways of microorganisms. It has many
advantages in surface analysis and has been used extensively to assess
marginal adaptation of restorations (National Institute of Health, U.S.,
2008). SEM has a large depth of focus and offers observations at low or high
magnification with a wide range of possibilities (Roulet et al., 1989).
It is well known that SEM requires dried samples and a complete
dehydration of specimens is necessary. With teeth specimen, because of the
intrinsic water content of dentin, a dehydration process must be performed
for adequate preparation for SEM. It has been shown that a gap can appear
wider after dehydration of dental tissues (Davila et al., 1989). Therefore, the
use of positive replicas instead of dehydrated samples is an established and
accepted method when marginal adaptation is analyzed with SEM (Taylor
and Lynch, 1993). The replica technique allows for observation of specimens
obtained in vitro and in vivo (Roulet et al., 1989). Replicas must be viewed in
a parallel plane in order to obtain reliable measurements. A small angulation
change may result in a marginal gap appearing smaller than it actually is
(Taylor and Lynch, 1993).
48
Evaluation of Marginal Adaptation Using SEM
As opposed to microleakage studies assessing dye penetration through
qualitative measurements, marginal adaptation has commonly been studied
and scored with quantitative margin analysis.
Quantitative Margin Analysis

Quantitative margin analysis has been described in 1984 by Porte et
al. (1984) and later in 1989 by Roulet et al. (1989), and Blunck et al. (1989).
It is currently the most widely used measurement in marginal adaptation
studies. In fact, this measurement is a quantitative assessment of different
qualitative criteria. Marginal quality of the entire studied margin is
described by different marginal quality criteria and, for each criteria, the
length of the marginal defect is measured on the SEM micrograph.
Percentages of the length belonging to each category are then calculated
(Taylor and Lynch, 1993). Even with the subjectivity resulting from
qualitative criteria, this method has been shown to be reliable in intraexaminer reliability tests (Roulet et al., 1989; Taylor and Lynch, 1993;
Sabatini, 2007).
Quantitative margin analysis has been described using different
criteria from simple to more complex. Frankenberger et al. (2008) based their
marginal adaptation criteria on a pass or fail basis where the categories were
continuous margin, gap/irregularity and not judgeable/artefact. Sabatini
et al. (Sabatini, 2007; Sabatini et al., 2010) identified seven marginal quality
49
criteria ranging from a perfect margin to gap width of more than 5
microns. The authors further reduced the number of categories into four
criteria for statistical purposes, and then furthermore into two categories
based on a pass or fail basis (gap or no gap). An analysis based on pass or fail
criteria can be misleading because it does not differentiate marginal defects,
for instance a marginal irregularity from a marginal gap. The clinical impact
of an irregularity versus a gap may be quite different for the restoration
longevity. On the other hand, too many criteria can lead to reliability issues.
Blunck et al. (Blunck and Roulet, 1989; Blunck and Zaslansky, 2011)
based their marginal adaptation assessment on four categories of marginal
quality (MQ): MQ1 is a continuous margin not or barely visible without or
with minor marginal irregularities without evidence of gap; MQ2 represents
severe marginal irregularities without evidence of gap; MQ3 corresponds to a
gap with a width of less than 2 microns without marginal irregularities; MQ4
is a gap wider than 2 microns with minor or severe marginal irregularities.
Summary
Finishing and polishing of resin-based composites are important for
esthetics and longevity of restorations. However, it has been shown and
discussed in this chapter that these procedures may disrupt the marginal
seal. Most of the literature on the effect of finishing and polishing techniques
on marginal adaptation evaluates restorations having a 90 cavosurface
angle. There is no information in the literature regarding the effect of
50
finishing and polishing direction (from the tooth structure to the resin-based
composite and from the resin-based composite to the tooth structure) on the
marginal
adaptation.
Several
questions
arise
and
need
further
investigations. Does the orientation of finishing/polishing technique affect

the marginal adaptation of resin-based composite restorations? Is the
marginal adaptation affected by the polishing system used? Is the marginal
adaptation influenced by the type of resin-based composite used (microhybrid
vs. nanofilled)? The present study aimed to answer these questions.
51
CHAPTER III
MATERIALS AND METHODS
Forty human molars extracted less than a year prior to the beginning
of the study with no visible enamel defect on the buccal and the lingual
surfaces were collected. All teeth were stored in a solution of water and
thymol since their extractions. Thymol was used to prevent bacterial growth
without affecting hard tissues of extracted teeth.
General Design
In order to test the effect finishing and polishing direction on marginal
adaptation, specimens were randomly divided in groups according to the
resin-based composite used, the type of finishing and polishing procedure
used and the direction of finishing and polishing procedures.
Specimens Preparation
Calculus and soft tissues were gently removed using 1.5 carbon steel
razor blades (Ted Pella Inc., USA) and a scaler (SH5/33, Hu-Friedy, USA).
Teeth were placed in distilled water in an ultrasonic cleaner (Cole-Palmer,
Chicago, IL, USA) for 5 minutes and then cleaned with a toothbrush (Oral-B,
USA) under running water.
52
Teeth Sectioning
Teeth were mounted and stabilized with sticky wax on a one square
inch plastic block. They were sectioned along their mesio-distal axis using a
precision saw machine (Isomet 1000, Buehler, Lake Bluff, IL, USA) with a
weight of 200g at a speed of 150 rpm with water coolant (Figures 5 and 6).
Dental pulp tissue was removed using a scaler (SH5/33, Hu-Friedy, USA).
Teeth were stored in a solution of chloramine T and artificial saliva (5000
ppm) at 4C. The solution was changed every 2 weeks. Each specimen
represents half of a whole a tooth.
Figure 5. Human molar sectioned along its mesio-distal axis using a precision
saw machine (Isomet 1000, Buehler, Lake Bluff, IL, USA).
53
Figure 6. Sections of a molar before preparation and restoration.
Enamel Preparation
Enamel was flattened using silicon carbide grinding paper of 200 mm
in diameter (Struers, Cleveland, OH, USA). Abrasive paper was positioned
on a MD-Fuga (Struers, Cleveland, OH, USA) and placed on the magnetic
fixation of a polishing machine (Rotopol-V, Struers, Cleveland OH, USA).
Abrasive paper of 320-grit, 600-grit and 1200-grit (Struers, Cleveland, OH,
USA) were used under water coolant at 70 rpm (Figure 7). Abrasive paper of
320-grit was used for 1 minute whereas 600- and 1200-grit papers were each
used for 30 seconds timed with a chronometer (Traceable timer, Control
Company, Friendswood, TX, USA). Careful attention was made to avoid
dentin exposure (Figure 8). Between each step, specimens were kept in
artificial saliva (pH: 6.99) to prevent dehydration.
54
Figure 7. Enamel was flattened using silicon carbide grinding paper on a

polishing machine (Rotopol-V, Struers, Cleveland OH, USA).
Figure 8. Flattened enamel without dentin exposure.
Standardized Preparation
A standardized triangular shape preparation was designed in order to
obtain two 30 bevels on each specimen. A distance of 3 mm between the
55
margins was arbitrarily chosen and calculations were made to determine the
required depth (Figure 9).
1,5 mm
3 mm
30
60
Tan 30 = X
1.5
X = 0.87 mm
Figure 9. Standardization of preparation in order to obtain two 30 bevels on

each specimen. Having set a width of 3 mm, calculations were made to
determine the exact depth of 0.87 mm.
Two 30 bevels were then prepared on the flattened enamel of each

specimen using a flat-end fine diamond bur (8847KR.31.016 Brasseler, USA)
with water coolant (Figure 10). The preparation had a triangular shape with
a depth of 0.87 mm and a width of 3 mm. Width and depth were precisely
measured with a 4 digital caliper (Model 47256, CenTech, Harbor Freight
Tools, Camarilla, CA, USA) (Figure 11). The end tip of the digital caliper was
modified to a triangular shape using a diamond bur to allow depth
measurement of the triangular preparation.
56
Figure 10. Bevel preparation using a flat-end

(8847KR.31.016 Brasseler, USA) with water coolant.
fine
diamond
bur
Figures 11. Preparations width and depth measured with a digital caliper
(Model 47256, Cen-Tech, Harbor Freight Tools, Camarilla, CA, USA).
57
The flattened enamel allowed for a better definition and identification
of the margins when observed with a field emission scanning electron
microscope (FeSEM) (Figures 12 and 13)
Figures 12. FeSEM micrographs of samples prepared without enamel

flattening. The outline of the preparation is curved and undefined.
58
A
Figures 13. FeSEM micrographs of samples prepared with enamel flattening.

The outline of the preparation is straight and well defined.
Specimens Restoration
All the steps were done by the same operator who rehearsed prior to
the beginning of the study in order to obtain a constant pressure and speed
during the finishing and polishing steps.
59
Groups
From the 40 extracted teeth, 80 specimens were obtained after teeth
sectioning and randomly distributed in different groups according to the type
of resin-based composite, the finishing and polishing system used and the
direction of the finishing and polishing procedure (Figure 14).
Figure 14. Group assignment and number of specimen according to the type
of resin-based composite, the finishing and polishing procedure and the
direction.
On each specimen, both margins were finished and polished with the
same technique, one from the resin-based composite to the tooth structure
(C-T) and the other from the tooth structure to the resin-based composite (TC) (Figure 15). Specimens were randomly divided into 4 groups. A list of
60
specimens placed in random group order was established using a random
sequence generator (Random.org, Dr. Mads Haahr, School of Computer
Science and Statistics, Trinity College, Dublin, Ireland) in order to randomly
restore, finish and polish one specimen in each group before moving forward
to the second specimen.
Figure 15. On each specimen, both margins were finished and polished with
the same technique, one from the resin-based composite to the tooth structure
(C-T) (left margin) and the other from the tooth structure to the resin-based
composite (T-C) (right margin).
Adhesion to Enamel
Specimens were etched with 35% phosphoric acid gel (Figure 16a)
(Ultraetch, Ultradent, South Jordan, UT USA) for 20 seconds and rinsed for
20 seconds with air and water spray. The time for each step was precisely
monitored with a chronometer (Traceable timer, Control Company,
61
Friendswood, TX, USA). Excess water was gently removed using paper towel
(Kimwipes EX-L, Roswell, GA, USA) to obtain a slightly moist surface.
OptiBond FL (Kerr, Orange, CA, USA) was used for the adhesive
procedure. The primer was applied with a microbrush applicator (Figure 16b)
(Kerr Applicators, Kerr, Orange, CA, USA) for 30 seconds and the solvent
was gently removed with air. The adhesive resin was then applied with a
new microbrush applicator in a thin layer (Figure 16c). Excess adhesive was
removed with a microbrush applicator wiped with a dry gauze. The adhesive
resin was light-cured for 20 seconds (Optilux 500, Demetron, Kerr, Danbury,
CT, USA) (Figure 16d).
a)
b)
c)
d)
Figures 16. Adhesion procedure using OptiBond FL.

a) Enamel was etched with 35 % phosphoric acid and rinsed.
b) After removal of excess water, primer was applied
c) Adhesive resin was applied in a thin layer.
d) Adhesive resin was light-cured for 20 seconds.
62
Restoration
Restoration Placement
Resin-based composite shade A1 was applied in one increment and
carefully adapted to minimize excess using an IPC instrument (Brasseler,
USA) and a brush (2 ROYAL Soft-Grip SG3010, China) (Figure 17). For the
Filtek Supreme Ultra (3M ESPE, St-Paul, MN, USA) group, the resin-based
composite was extruded from the unit-dose container directly into the
preparation. For the Point 4 (Kerr, Orange, CA, USA) group, the resin-based
composite was extruded from a syringe. The amount needed was cut from the
syringe, carried with an IPC and placed into the preparation. The resinbased composite was light-cured for 40 seconds with the tip of the light held
as close as possible to the surface without contacting it. The same curing
light (Optilux 500, Demetron, Kerr, Danbury, CT, USA) was used for all
polymerization steps. The intensity of the curing light was verified
periodically using the radiometer on the unit to ensure that at least 600
mW/cm2 was delivered to the restorative material.
Finishing and Polishing

Restorations were finished and polished immediately after their
polymerization. The direction of finishing and polishing was dictated by the
movement of the abrasive particles on the finishing and polishing instrument
(either from T-C or C-T) rather than solely by the hand movement because
the speed of the bur is much faster than the hand movement.
63
Figure 17. Placement of resin-based composite in one increment with careful

adaptation to minimize excess using an IPC instrument (Brasseler, USA) and
a brush (2 ROYAL Soft-Grip SG3010, China)
Sof-Lex
For the Sof-Lex XT (3M ESPE, St-Paul, MN, USA) groups, finishing
was performed without water using the red and dark orange Sof-Lex XT
discs on a slow-speed handpiece with a speed reducer. Polishing was then
executed using the light orange and yellow discs without a speed reducer
(Figure 18).
Diamond Bur, Sof-Lex disc and Rubber Polishers

In the second finishing and polishing group, finishing was completed
without water with a diamond bur on a highspeed handpiece (Figure 19a)
followed by a dark orange Sof-Lex XT disc with a speed reducer (Figure 19b).
Polishing followed with HiLuster Plus rubber polishing cups (HiLuster,
Kerr, Bioggio, Switzerland) on a slow speed handpiece with a speed reducer.
64
A blue rubber polishing cup was used first with water (Figure 19c), then the
surface was rinsed and dried and a grey rubber polishing cup was used
(Figure 19d).
Figure 18. Sof-Lex group. Finishing and polishing done with a series of SofLex XT discs (3M ESPE, St-Paul, MN, USA).
a)
b)
c)
d)
Figures 19. Second polishing technique.

a) Finishing initiated with a fine diamond finishing bur
b) Finishing continued with a dark orange Sof-Lex disc
c) Polishing initiated with a blue rubber polishing cup with water
d) Polishing completed with a dry grey rubber polishing cup
65
Verification of Polishing and Margins
Specimens were observed under a light microscope at 20X (Zeiss,
Thornwood, NY, USA) to verify the quality of the polished surface (Figure
20). Polishing was corrected as needed.
Figure 20. Light microscope (Zeiss, Thornwood, NY, USA) used to verify the
quality of the polishing surface
Specimens were stored in artificial saliva no longer than 4 hours

before taking impression for replicas. This time limit was set in order to
minimize resin-based composite swelling by hygroscopic expansion.
66
Specimens Preparation for FeSEM Observation
Fabrication of Replicas
Specimens were placed in 70% ethanol (Figure 21a) in an ultrasonic
bath for 2 minutes (Figure 21b and 21c). Specimens were then rinsed with
air and water spray and air dried to remove excess moisture that could
interfere with the impression process. Impressions were immediately taken
using a low viscosity polyvinyl siloxane (Aquasil XLV Ultra, fast set,
Dentsply Caulk, Milford, DE, USA). A first impression was taken and let to
set for 5 minutes in order to further remove any contaminant from the
surface, especially from the tooth-restoration interface. The material was
then peeled from the surface of the specimen and the impression was
disposed. A second impression using the same light viscosity material was
taken and left to set for 10 minutes (Figure 21d). The specimen was carefully
removed from the impression and stored in artificial saliva at 4C. The
impression was inspected for any imperfection (Figure 21e) and placed on a
double sided tape in a sealed plastic container. Because gas formed during
the polymerization of the impression material can escape from the material
after it sets, the impression was allowed to rest for 24 hours at room
temperature.
Impressions were subsequently poured with low viscosity epoxy resin
(Epoxicure, Buehler Ltd, Lake Bluff, IL, USA) to create positive replicas. The
epoxy resin was mixed in a 5:1 ratio. Therefore, five parts of Epoxycure resin
(no 20-8130-032) was mixed with one part of Epoxide Hardener (no 20-8132-
67
008) and placed in a vacuum for 10 minutes to allow air bubbles to escape
from the mixture. A drop of epoxy resin was then poured in the impression
with a plastic pipette and gently spread uniformly with air to improve its
adaptation into the impression. The impression was then completely filled
with epoxy resin and allowed to set undisturbed for 24 hours in a fume hood.
a)
b)
d)
c)
e)
Figure 21. Specimens preparation for FeSEM observations.

a) Specimens placed in 70% ethanol solution
b) Ultrasonic cleaner
c) Specimens placed in the ultrasonic bath for 2 minutes
d) Impressions taken with a low viscosity vinyl polysiloxane
e) Impression inspected for any imperfection.
68
Mounting of Replicas, Sputter Coating
Replicas were mounted on aluminum stubs (Ted Pella, Inc., USA).
Aluminum stubs were first roughened with silicone carbide paper 320-grit
and placed in 70% ethanol in an ultrasonic bath for 2 minutes.
Replicas were removed from the impression and mounted on
aluminum stubs using colloidal graphite paint (Ladd Research Industries
Inc., Burlington, VT, USA). Care was taken to ensure that the margins on
the replicas would be parallel to the stub base. When needed, a double-sided
carbon tape (NEM Tape, Nisshin EM.CO, LTD) was used to stabilize the
replicas on the stub prior to the application of colloidal graphite paint. The
colloidal graphite paint was allowed to dry for at least 6 hours.
Using a caliper (Staedler, Nuernberg, Germany), a length of 2 mm was
outlined on the margins and only that portion was analyzed. A vertical line
was made with a #15 blade to separate the two margins and to facilitate the
localization of the margins under the scanning electron microscope.
Replicas were sputter coated with gold and palladium (Emitech K550,
Ashford, Kent, UK) at 10 mA for 2.5 minutes (Figures 22).
FeSEM Observation
Replicas were observed using a field emission scanning electron
microscope (FeSEM) (Hitachi S-4800, Hitachi High Technologies America,
Inc. Pleasanton, CA, USA) at 3.0 kV and 10uA.
69
Figures 22. Sputter coating of samples prior to observation under FeSEM.
Initially, replicas were observed at a 30X magnification to localize

enamel-restoration interfaces and then at 200X magnification to capture
images of the designated portion of the margin (Figures 23).
Figures 23. Initial observation of replicas at low magnification to localize the

margins. These two images also show the 2 mm outline marks done with a
caliper prior to sputter coating and the vertical central mark intended to
separate and localize the two margins.
70
Five to seven images were captured for each portion of the margin and
saved in a tiff format.
Images Preparation
In order to assess the marginal adaptation, images taken for a same
margin were merged using Photoshop Elements 3.0 (Adobe Systems Inc., San
Jose, CA, USA). Except for the bottom picture of the margin, images were
cropped to remove the scale bar and obtain an optimal merged panorama
with a continuous margin. The canvas size of the merged panorama was
slightly enlarged and converted to a black background to increase the
contrast. These merged panoramas were also saved in a tiff format. Images of
margins were assigned to a random numerical code in order to blind the
examiner to the experimental groups during measurements.
Evaluation of Marginal Quality

The marginal adaptation was measured using ImageJ software
(ImageJ 1.44p, Wayne Rasband, National Institute of Health, USA). Merged
images were observed at a size of 100%. First, a straight line was drawn as a
reference next to the margins and all measurements were done linearly and
parallel to that reference line for optimal accuracy. The length of the entire
studied margin (approximately 2 mm) was measured exactly. The length of
any artifact such as bubbles or contamination was also measured and
subsequently subtracted from the total length of the studied margin. All
71
defects belonging to each marginal quality criteria were then measured and
saved in an excel file where calculations were done (Figure 24).
Figure 24. Marginal adaptation measured with ImageJ software (ImageJ

1.44p, Wayne Rasband, National Institute of Health, USA). The length of the
entire studied margin was measured as well as the length of any artifact and
all defects belonging to each marginal quality criteria.
The quality of the margin was expressed as a percentage of the entire

studied margin belonging to each marginal quality criterion previously
defined by Blunck et al. (Blunck and Roulet, 1989; Blunck and Zaslansky,
2011). Marginal irregularity criterion (MQ2) was adapted for the present
study (Table 1 and Figures 25 to 28).
72
Intra-Observer Reliability
Intra-observer reliability tests were performed to verify the reliability
of the quantitative measurements made by a single observer for each
marginal quality criteria. All measurements (entire studied margin length,
artifacts and marginal defects) were taken twice by a same blinded observer
in a two-week interval for all specimens included in the pilot study, for a
total of 84 specimens.
Table 1. Marginal Quality Criteria

Marginal Quality*
Definition
MQ1
Margin not or hardly visible

No or slight marginal irregularities*
No gap
MQ2
No gap but severe marginal irregularities
MQ3
Gap visible (hairline crack up to 2 m)
MQ4
Severe gap (more than 2 m)

Slight and severe marginal irregularities
*Marginal irregularities means porosity, hairline defect (no gap), roughness

in the resin-based composite.
Source: Adapted from Blunck U, Zaslansky P. (2011). Enamel margin
integrity of Class I one-bottle all-in-one

restorations. J Adhes Dent 13(1):23-29.
adhesives-based
73
A
Figures 25. Marginal quality 1 criterion (MQ1).

MQ1 is represented by a margin barely visible (A) and a margin clearly
visible (B) without any defect or presenting slight marginal irregularities.
Images on the left are several FeSEM images merged together using
Photoshop Elements 3.0.
74
A

MQ2 is represented by a hairline defect (A) and roughness at the toothrestoration interface (B) without any gap. Images on the left are several SEM
images merged together using Photoshop Elements 3.0.
75
A

MQ3 is represented by a hairline crack of less than 2 m. Image on the left
are several SEM images merged together using Photoshop Elements 3.0.
76

MQ4 is represented by a severe gap (more than 2 m) and severe marginal
irregularities. Images on the left are several SEM images merged together
using Photoshop Elements 3.0.
77
Statistical Methods
Variables
In this study, independent variables are the type of resin-based
composite (FSU and PT4), the finishing and polishing technique (SL and R),
and the polishing direction (C-T and T-C), while dependent variables are
continuous margins (MQ1), marginal irregularities (MQ2) and gaps
(MQ3+MQ4). For statistical purposes and to simplify the results obtained,
marginal quality MQ3 and MQ4 (both representing gaps) were combined in
one dependent variable by taking the sum of the two variables.
Power Analysis
To estimate the sample size required for the present study, a pilot
study (n=6 per group) was conducted. This analysis suggested that a sample
size of 20 specimens per group was needed to achieve 80% power at two-sided
5% significance level and to be able to detect a difference as small as 7.50 in
MQ1 and as large as a common standard deviation of 25.
Statistical Analysis
Intra-class
correlation
coefficient
was
used
for
intra-observer
reliability of the measurement technique. Moreover, a paired-sample t-test

was used to assess if there was a significant difference between the two
measurements made by a single observer.
78
Univariate analysis was performed to explore each variable in the
study. Since two polishing directions were on the same specimen, a pairedsample t-test and the nonparametric Wilcoxon signed-rank test (i.e. when
data were not normally distributed) were used. A two-sample t-test and the
nonparametric Wilcoxon rank-sum test (i.e. data were not normally
distributed) were used to detect the difference between two polishing
techniques and two resin-based composites.
One-way ANOVA with the post-hoc Tukeys HSD test was performed
to test for a difference among all eight experimental groups. If the data were
not normally distributed, rank transformation was conducted. Subsequently,
one-way ANOVA based on ranked data, an equivalent test statistic to the
nonparametric Kruskal-Wallis test, followed by post-hoc Bonferroni test was
carried out.
To evaluate the interaction between resin-based composites and
polishing techniques on MQ1, MQ2 and MQ3+MQ4, two-way ANOVA was
performed. When the data were not normally distributed, two-way ANOVA
based on ranked data was performed. Additionally, when taking into account
the correlated data obtained from the same specimen, correlations among
measurements made on the same specimen were modeled using the random
effects in Mixed Model ANOVA (i.e. to allow correlation between two
measurements of polishing directions obtained from the same specimen) to
evaluate the effects of resin-based composite, polishing technique, polishing
79
direction on MQ1, MQ2, and MQ3+MQ4, including interactions among the
three factors.
The Shapiro-Wilk test was used to check for the normality of the
variables.
A p-value of less than 0.05 was used as a criterion for statistical
significance. SAS for Windows (v9.2, SAS Institute Inc, Cary, NC, USA) was
used for the data analysis.
Summary
Forty human molars were sectioned along their mesio-distal axis.
Buccal and lingual enamel was flattened and a triangular preparation
(0.87mm deep and 3mm wide) forming two 30 bevels was achieved.
Specimens were randomly assigned in groups (n=20) and restored with two
resin-based composites: a nanofilled (Filtek Supreme Ultra) and a
microhybrid (Point4) and two finishing/polishing techniques: a series of SofLex discs and a sequence of diamond bur followed by a dark-orange Sof-Lex
disc and rubber polishing cups. On each specimen, both margins were
finished and polished with the same technique, one from resin-based
composite to tooth structure and the other from tooth structure to resinbased composite. Replicas were fabricated and sputter coated with goldpalladium for FeSEM observation (200X). Images were merged together and
quantitative margin analysis was performed based on four defined marginal
quality criteria. To detect differences between the two polishing directions,
80
data was analyzed using a paired-samples t-test and the nonparametric
Wilcoxon signed-rank test (when data were not normally distributed). A twosample t-test and the nonparametric Wilcoxon rank-sum test were used to
detect differences between the two polishing techniques and the two resinbased composites. One-way ANOVA with the post-hoc Tukeys HSD test or
Bonferroni test were performed to test for a difference among the eight
experimental groups. Two-way ANOVA was performed to evaluate the main
effects of the independent variables and to test the interactions between
them.
Table 2. Resin-Based Composite Specifications

Resin-based
Composite
Shade
Resin
Filler
Wt%
Vol%
Size
Shape
Lot number
Filtek
Supreme
Ultra
(nanofilled)
A1B
Bis-GMA
UDMA
TEGDMA
PEGDMA
Bis-EMA
Zirconia
Silica
Zirconia/Silica
nanoclusters
72.5
55.6
20 nm (silica)
4 to 11 nm
(zirconia)
0.6 to 1.0
microns
(clusters)
Spherical
N202525
Point 4
(microhybrid)
A1B
Bis-GMA
Barium glass
Aluminoborosilicate
glass
Silicon dioxide
76
57
Average of 0.4
microns
Irregular
3312262
Source: 3M ESPE, 2009, Filtek Supreme Ultra Universal Restorative. Available at:
http://www.everyonehasashade.com/main/pdf/IFU_supreme_ultra_v3.pdf. Accessed July, 2011.
Kerr Dental Sybron Dental Specialties. A Guide for using Point 4 Optomized Particle Composite System.
Available
at:
http://www.kerrdental.com/index/cms-filesystem-action?file=/kerrdental-products-dfu/dfupoint4.pdf. Accessed July, 2011.
Watanabe H, Khera SC, Vargas MA, Qian F., 2008. Fracture toughness comparison of six resin composites.
Dent Mater 24(3):418-425.
81
Table 3. Adhesive Specifications

Adhesive
Category
Filler
% Filler
Size
Lot number
Primer
Lot number
Adhesive
OptiBond FL
3-step totaletch
Barium glass
48
0.6 micron
3448800
3525628
Source: Kerr. Instruction For Use OptiBond FL (product guide)
Table 4. Sof-Lex Finishing and Polishing Discs Specifications

Instrument
Abrasive
Abrasive Particles Size
Sof-Lex XT Red
Aluminum oxide
40 m*
Sof-Lex XT Dark Orange
Aluminum oxide
24 m*
Sof-Lex XT Light Orange
Aluminum oxide
20-30 m*
Sof-Lex XT Yellow
Aluminum oxide
8 m*
Source: 3M ESPE, Sof-Lex Technical Product Profile, http://multimedia.3m.com
82
*Jefferies SR. (1998). The art and science of abrasive finishing and polishing in restorative dentistry. Dent
Clin North Am 42(4):613-627.
Table 5. HiLuster Rubber Polishers Specifications

Instrument
HiLuster Gloss Plus Polisher
(blue)
HiLuster Plus Dia Polisher
(gray)
Source:
Abrasive
Abrasive Particle Size
Aluminum oxide
8 m*
Diamond
5 m*
Kerr Dental, 2011, HiLuster (Plus) Systme de Polissage, http://www.kerrdental.fr/finition-etpolissage/polissage-des-composites/productfamily/HiLusterPolishingSystem

* Reality Esthetics, 2003, Hawe HiLuster specifications,
http://www.realityesthetics.com/protected/book/2003/Ratings/Rubber_Polish.pdf
83
84
CHAPTER IV
RESULTS
Intra-Observer Reliability Measurements

Eighty-four margins, representing all the specimens included in the
pilot study, were tested for intra-observer reliability of the measurement
technique. Intraclass correlation coefficient was computed to assess intraobserver agreement in tracing of marginal defects. Overall, there was very
strong evidence that the intraclass correlation differed from zero (p<0.0001
for each instance), and the correlation coefficients of 0.99 for MQ1, 0.99 for
MQ2, 0.97 for MQ3, and 0.99 for MQ4 indicated a strong agreement between
two measurements made by the single observer.
Additional analysis was conducted to determine if there was a
significant difference between two measurements using a paired-sample ttest. The data revealed that there was no statistically significant difference
between two measurements for MQ1 (p=0.1085), for MQ2 (p=0.1449), for
MQ3 (p=0.4240) and for MQ4 (p=0.1430).
Descriptive Findings for Marginal Quality Criteria

The mean values for each marginal quality criteria (MQ1, MQ2 and
MQ3+MQ4) were calculated individually for all study groups along with their
standard deviations, medians, maximums and minimums. These values were
85
expressed in percentage of the entire studied margin and are presented in
Figure 28 and in Table 6.
Figure 29. Mean values with the standard deviation for continuous margin
(MQ1), marginal irregularities (MQ2) and gaps (MQ3+MQ4) among the
eight experimental groups expressed in % of entire margin length studied.
86
Table 6. Descriptive Statistics of Marginal Quality by Treatment Groups
GROUPS
Variable
Mean
(%)
SD
Min
Max
Median
FSU / SL / C-T
MQ1
MQ2
MQ3+MQ4
20
20
20
91.38
7.62
1.01
11.02
10.58
2.69
51.56
0.00
0.00
100.00
48.44
11.34
94.39
4.95
0.00
FSU / SL / T-C
MQ1
MQ2
MQ3+MQ4
20
20
20
80.91
14.89
4.21
11.97
12.54
7.42
54.29
0.00
0.00
100.00
45.71
23.49
81.71
13.95
0.13
FSU / R / C-T
MQ1
MQ2
MQ3+MQ4
20
20
20
97.74
2.26
0.00
4.11
4.11
0.00
84.73
0.00
0.00
100.00
15.27
0.00
100.00
0.00
0.00
FSU / R / T-C
MQ1
MQ2
MQ3+MQ4
20
20
20
86.87
8.66
4.48
12.30
8.87
7.50
58.53
0.00
0.00
100.00
33.16
29.87
90.86
5.27
0.00
PT4 / SL / C-T
MQ1
MQ2
MQ3+MQ4
20
20
20
96.96
2.99
0.04
5.17
4.99
0.20
77.63
0.00
0.00
100.00
21.48
0.89
98.34
1.66
0.00
PT4 / SL / T-C
MQ1
MQ2
MQ3+MQ4
20
20
20
76.74
16.98
6.28
11.98
12.77
8.89
59.11
0.00
0.00
100.00
40.89
34.53
74.41
14.19
1.46
PT4 / R / C-T
MQ1
MQ2
MQ3+MQ4
20
20
20
98.51
1.45
0.04
1.94
1.83
0.16
92.76
0.00
0.00
100.00
6.53
0.71
99.52
0.48
0.00
PT4 / SL / T-C
MQ1
MQ2
MQ3+MQ4
20
20
20
76.26
14.16
9.57
13.79
10.89
12.02
52.48
0.00
0.00
100.00
34.60
44.89
78.15
11.61
5.78
87
Difference between Two Polishing Directions
Results of the paired-sample t-test for MQ1 and MQ2 and the
Wilcoxon signed-rank test for MQ3+MQ4 (data were not normally
distributed) revealed significant differences between the two polishing
directions C-T and T-C for each marginal quality criteria (p<0.05).
Data showed systematically more continuous margins (MQ1) and
less marginal irregularities (MQ2) when the polishing procedures were
directed from C-T than from T-C. Regarding gaps (MQ3+MQ4), FSU/SL
showed marginally significant difference between the two directions
(p=0.0537). Less gaps were found with the C-T direction in all the other
groups (p<0.05). These values are presented in Tables 6 and 7 and described
in details below.
Continuous Margins (MQ1)

Results of a paired-sample t-test showed a significant difference in
MQ1 between the two polishing directions within all combinations of resinbased composite and finishing/polishing techniques: FSU/SL (p=0.0055),
FSU/R (p=0.0016), PT4/SL (p<0.0001) and PT4/R (p<0.0001). The data
showed that the mean MQ1 for polishing direction C-T was significantly
greater than that of polishing direction T-C (Tables 6 and 7).
88
Marginal Irregularities (MQ2)
Based on a paired-sample t-test, there was a significant difference in
MQ2 between the two polishing directions within all combinations of resinbased composites and finishing/polishing techniques: FSU/SL (p=0.0388),
FSU/R (p=0.0085), PT4/SL (p=0.0002) and PT4/R (p<0.0001). The data
showed that the mean MQ2 for polishing direction T-C was significantly
greater than that of polishing direction C-T (Tables 6 and 7).
Gaps (MQ3+MQ4)
In this category, the data were not normally distributed and Wilcoxon
signed-rank test was used. There was no significant difference in MQ3+MQ4
between the two polishing directions within FSU/SL (p=0.0537).
However, there was a significant difference in MQ3+MQ4 between the
two polishing directions within FSU/R (p=0.0039), PT4/SL (p=0.0005) and
PT4/R (p<0.0001). The data showed that the mean MQ3+MQ4 for polishing
direction T-C was significantly greater than that of polishing direction C-T
(Tables 6 and 7).
Table 7. Comparison in Marginal Quality Difference between Two Polishing Directions under Different Conditions
GROUPS
Variables
Mean
(%)
SD
Min
Max
Median
FSU / SL
Difference in Direction for MQ1
Difference in Direction for MQ3+MQ4
20
20
20
10.47
-7.27
-3.20
14.95
14.65
7.75
-18.38
-45.71
-20.07
45.71
22.44
11.34
13.64
-9.19
0.00
0.0055*
0.0388*
0.0537
FSU / R
20
20
20
10.87
-6.40
-4.48
13.27
9.76
7.50
-15.27
-33.16
-29.87
41.47
15.27
0.00
6.57
-5.17
0.00
0.0016*
0.0085*
0.0039**
PT4 / SL
20
20
20
20.22
-13.99
-6.23
13.46
13.78
8.89
-11.58
-36.43
-34.53
40.05
18.72
0.00
20.74
-12.07
-1.46
<0.0001*
0.0002*
0.0005**
PT4 / R
20
20
20
22.25
-12.71
-9.54
13.86
10.87
12.05
-2.08
-31.82
-44.89
45.49
2.08
0.71
19.94
-9.45
-5.78
<0.00001*
<0.0001*
<0.0001**
p-value
Significantly different at p<0.05, with a paired-sample t-test
** Significantly different at p<0.05, with a Wilcoxon signed-rank test

Note: Direction change = polishing direction from resin-based composite to tooth structure (C-T) polishing
direction from tooth structure to resin-based composite (T-C)
89
90
Difference between Two Polishing Techniques
A two-sample t-test and the Wilcoxon rank-sum test (when the data
were not normally distributed) were used to assess the difference in marginal
adaptation between the two polishing techniques.
For FSU/T-C, PT4/C-T and PT4/T-C groups, there was no significant
difference between the two polishing techniques (SL vs. R) (p>0.05).
However, in the FSU/C-T group, results revealed more continuous margins
(MQ1) (p=0.0036), less marginal irregularities (MQ2) (p=0.0051) and less
gaps (MQ3+MQ4) (p=0.0402) when using rubber polishers (R) compared to
Sof-Lex discs (SL). These values are presented in Tables 6 and 8 and
described in details below.

Results revealed a significant difference in MQ1 between the two
polishing techniques for FSU/C-T (p=0.0036, Wilcoxon rank-sum test). The
data showed that the mean MQ1 for R was significantly greater than for SL
(Tables 6 and 8).
However, there was no significant difference in MQ1 between the two
polishing techniques for FSU/T-C (p=0.0988, Wilcoxon rank-sum test),
PT4/C-T (p=0.5012, Wilcoxon rank-sum test) and PT4/T-C (p=0.9078, twosample t-test) (Tables 6 and 8).
91
Based on Wilcoxon rank-sum test, there was a significant difference in
MQ2 between the two polishing techniques for FSU/C-T (p=0.0051). The data
showed that the mean MQ2 for SL was significantly greater than R (Tables 6
and 8).
polishing techniques for FSU/T-C (p=0.1225, Wilcoxon rank-sum test),
PT4/C-T (p=0.4654, Wilcoxon rank-sum test) and PT4/T-C (p=0.4572, twosample t-test) (Tables 6 and 8).
Gaps (MQ3+MQ4)
Results of Wilcoxon rank-sum test demonstrated a significant
difference in MQ3+MQ4 between the two polishing techniques for FSU/C-T
(p=0.0402). The data showed that the mean MQ3+MQ4 for SL was
significantly greater than for R (Tables 6 and 8).
However,
no
significant
differences
were
found
for
FSU/T-C
(p=0.9999), PT4/C-T (p=0.9999) and PT4/T-C (p=0.1849) (Tables 6 and 8).
Table 8. Comparison in Marginal Quality Difference between Two Polishing Techniques under Different Conditions
GROUPS
Variables
Mean
(%)
SD
Min
Max
Median
p-value
FSU / C-T
Difference between SL and R for MQ1
Difference between SL and R for MQ3+MQ4
20
20
20
-6.36
5.36
1.01
11.31
11.05
2.69
-48.44
-5.01
0.00
5.01
48.44
11.34
-3.68
2.46
0.00
0.0039*
0.0051*
0.0402*
FSU / T-C
20
20
20
-5.96
6.23
0.27
13.11
11.92
9.41
-40.43
-13.99
-29.87
21.62
40.43
16.60
-4.06
3.09
0.00
0.0988
0.1225
0.9999
PT4 / C-T
20
20
20
-1.55
1.54
0.01
6.09
5.88
0.26
-22.37
-6.53
-0.71
7.24
21.48
0.89
0.00
0.00
0.00
0.5012
0.4654
0.9999
PT4 / T-C
20
20
20
0.48
2.82
-3.30
19.47
14.90
13.51
-37.67
-25.43
-31.41
29.64
28.79
31.19
5.05
2.13
-3.48
0.9078
0.4572
0.1849
* Significantly different at p<0.05, with the Wilcoxon rank-sum test

Note: Difference between SL and R = polishing technique SL polishing technique R
92
93
Difference between Two Types of
Resin-Based Composite
Two statistical tests were used to assess the difference between the
two resin-based composites. A two-sample t-test was used when the data
were normally distributed and a Wilcoxon rank-sum test was used when the
data did not present a normal distribution.
Results did not show much consistency. It seems that when polishing
with SL with C-T direction, PT4 showed more continuous margins (MQ1)
(p=0.0160, Wilcoxon rank-sum test) and less marginal irregularities (MQ2)
(p=0.0232, Wilcoxon rank-sum test) than FSU but no significant difference
was found for gaps (MQ3+MQ4) (p=0.1381 Wilcoxon rank-sum test).
However, when polishing with R from T-C direction, FSU showed more
continuous margins (MQ1) (p=0.0172, Wilcoxon rank-sum test) and less
gaps (MQ3+MQ4) (p=0.0381, Wilcoxon rank-sum test) than PT4 but no
difference was found for marginal irregularities (MQ2) (p=0.1439, Wilcoxon
rank-sum test). No significant differences were found in the other groups:
SL/T-C, R/C-T (p>0.05). These values are presented in Tables 6 and 9 and
described in details below.

Results of the Wilcoxon rank-sum test revealed a significant difference
in MQ1 between the two resin-based composites for SL/C-T (p=0.0160,
Wilcoxon rank-sum test) and R/T-C (p=0.0172, Wilcoxon rank-sum test). For
94
SL/C-T, the data showed that the mean MQ1 was greater for PT4 than for
FSU, whereas for R/T-C, the mean MQ1 for FSU was significantly greater
than PT4.
There was no significant difference in MQ1 between the two resinbased composites for SL/T-C (p=0.2782, two-sample t-test), R/C-T (p=0.7072,
Wilcoxon rank-sum test) (Tables 6 and 9).

There was a significant difference in MQ2 between the two resin-based
composites for SL/C-T (p=0.0232, Wilcoxon rank-sum test). Data showed that
the mean MQ2 for FSU was significantly greater than PT4.
resin-based composites for SL/T-C (p=0.2782, two-sample t-test), R/C-T
(p=0.7296, Wilcoxon rank-sum test) and R/T-C (p=0.1439, the Wilcoxon ranksum test) (Tables 6 and 9).
Gaps (MQ3+MQ4)
Based on the Wilcoxon rank-sum test, there was no significant
difference in MQ3+MQ4 between the two resin-based composites for SL/C-T
(p=0.1381), SL/T-C (p=0.4188) and R/C-T (p=0.3421). A significant difference
in MQ3+MQ4 was found for R/T-C (p=0.0381) where the mean MQ3+MQ4 for
PT4 was significantly greater than for FSU.
Table 9. Comparison in Marginal Quality between Two Resin-based Composites under Different Conditions
GROUPS
Variable
Mean
SD
Min
Max
Median
p-value
(%)
SL / C-T
Difference between FSU and PT4 for MQ1
Difference between FSU and PT4 for MQ3+MQ4
20
20
20
-5.59
4.62
0.96
11.93
11.42
2.70
-46.44
-12.18
0.00
11.73
46.44
11.34
-3.75
3.12
0.00
0.0160*
0.0232*
0.1381
SL / T-C
20
20
20
4.17
-2.10
-2.07
17.31
15.50
11.13
-22.61
-36.48
-34.53
36.48
21.57
18.13
2.21
-3.31
0.00
0.2782
0.6034
0.4188
R / C-T
20
20
20
-0.77
0.81
-0.04
4.67
4.61
0.16
-15.27
-6.53
-0.71
7.24
15.27
0.00
0.00
0.00
0.00
0.7072
0.7296
0.3421
R / T-C
20
20
20
10.61
-5.51
-5.10
22.04
13.65
15.25
-41.47
-27.78
-44.89
45.49
21.99
29.87
9.70
-4.30
-3.89
0.0172*
0.1439
0.0381*
* Significantly different at p<0.05, Wilcoxon rank-sum test

Note: Difference between FSU and PT4 = Resin-based composite FSU resin-based composite PT4
95
96
Difference between the Eight Experimental Groups
Under the assumption of independent samples, MQ1, MQ2 and
MQ3+MQ4 data were first analyzed using the ShapiroWilk test to verify the
assumption of normality. Since the data were not normally distributed, oneway ANOVA based on ranked data (i.e. equivalent to non-parametric
Kruskal-Wallis test) was performed for each dependent variable (MQ1, MQ2
and MQ3+MQ4).
This analysis gives an overall idea of the difference between the eight
experimental groups. More statistical analyses were performed to determine
the main effects and the possible interactions between the independent
variables.
Differences in Continuous Margins (MQ1)

Results revealed that there was a significant effect for the types of
experimental groups on MQ1 (p<0.0001). The post-hoc Bonferroni test
showed that the mean MQ1 for PT4/R/C-T and FSU/R/C-T groups was
significantly greater than the mean of all other groups except PT4/SL/C-T,
while no significant differences was found among PT4/R/C-T, FSU/R/C-T, and
PT4/SL/C-T, and between PT4/SL/C-T and FSU/SL/C-T, between FSU/R/T-C
and FSU/SL/T-C and among FSU/SL/T-C, PT4/SL/T-C and PT4/R/T-C. All
groups with a C-T direction had a significantly higher MQ1 mean than the
groups with a T-C direction. Table 10 reports more detailed results of the
post-hoc Bonferroni test.
97
Table 10. Mean MQ1 by Experimental Groups
Experimental
Groups
Mean MQ1
(mean ranking score)
(%)
Group
Comparison**
PT4 / R / C-T
20
98.51 (120.53)
FSU / R / C-T
20
97.74 (120.45)
PT4 / SL / C-T
20
96.96 (113.03)
FSU / SL / C-T
20
91.38 (86.10)
FSU / R / T-C
20
86.87 (71.78)
FSU / SL / T-C
20
80.91 (50.78)
PT4 / SL / T-C
20
76.74 (40.78)
PT4 / R / T-C
20
76.26 (40.58)
B
B
**means with the same letter are not significantly different using post-hoc
Bonferroni test (p>0.05).
Differences in Marginal Irregularities (MQ2)

Results revealed a significant effect for the types of experimental
groups on MQ2 (p<0.0001). The post-hoc Bonferroni test showed that the
mean MQ2 for PT4/SL/T-C, FSU/SL/T-C and PT4/R/T-C groups was
significantly greater than that for the other groups except for FSU/R/T-C and
FSU/SL/C-T. Table 11 reports more detailed results of the post-hoc
Bonferroni test.
98
Table 11. Mean MQ2 by Experimental Groups
Experimental
Groups
Mean MQ2
(%)
Group
Comparison**
PT4 / SL / T-C
20
16.98 (114.43)
FSU / SL / T-C
20
14.89 (108.08)
PT4 / R / T-C
20
14.16 (106.13)
FSU / R / T-C
20
8.66 (89.73)
FSU / SL / C-T
20
7.62 (82.15)
PT4 / SL / C-T
20
2.99 (53.63)
FSU / R / C-T
20
2.26 (45.73)
PT4 / R / C-T
20
1.45 (44.15)
Differences in Gaps (MQ3+MQ4)

Results revealed a significant effect for the types of experimental
groups on MQ3+MQ4 (p<0.0001). The post-hoc Bonferroni test showed that
the mean MQ3+MQ4 for PT4/R/T-C was significantly greater than that for
the other groups except for PT4/SL/T-C and FSU/R/T-C. Moreover, except for
FSU/SL/C-T, all groups with a C-T direction had a significantly lower
MQ3+MQ4 mean than the groups with a T-C direction. Table 12 reports
more detailed results of the post-hoc Bonferroni test.
99
Table 12. Mean MQ3+MQ4 by Experimental Groups
Experimental
Groups
Mean MQ3+MQ4
(%)
Group
Comparison**
PT4 / R / T-C
20
9.57 (123.53)
PT4 / SL / T-C
20
6.28 (103.40)
FSU / R / T-C
20
4.48 (92.15)
FSU / SL / T-C
20
4.21 (90.73)
FSU / SL / C-T
20
1.01 (68.05)
PT4 / SL / C-T
20
0.04 (56.38)
PT4 / R / C-T
20
0.04 (56.28)
FSU / R / C-T
20
0.00 (53.50)
Interaction between Resin-Based Composite

and Polishing Technique for Each
Polishing Direction
In order to test the interaction between the resin-based composite and
the polishing technique, two-way ANOVA was performed whenever the
assumption of normality was valid. When the data were not normally
distributed, two-way ANOVA based on ranked data was performed.
Comparisons were made for polishing directions C-T and T-C.
100
1. Interaction Within Polishing Direction C-T
Results of two-way ANOVA based on ranked data revealed a
statistically significant main effect for polishing technique for MQ1
(F(1,76)=8.92; p=0.0038) and MQ2 F(1,76)=8.47; p=0.0047). No significant
main effects for the polishing technique was found for MQ3+MQ4
(F(1,76)=3.18; p=0.0784). The post-hoc Bonferroni test indicated that the
mean MQ1 observed with R was significantly greater than that with SL
(mean MQ1: 98.133.20 vs. 94.178.95, respectively) and the mean MQ2
observed with SL was significantly greater than that with R (mean MQ2:
5.308.49 vs. 1.853.17, respectively).
The main effect for the type of resin-based composite proved to be
non-significant for MQ1 (F(1,76)=2.97, p=0.0887), MQ2 (F(1,76)=2.68,
p=0.1059) and MQ3+MQ4 ((F(1, 76)=0.91, p=0.3444).
The
interaction
between
resin-based
composite
and
polishing
technique also proved to be non-significant for MQ1 (F (1,76) =3.88;

p=0.0526), MQ2 (F (1,76) =3.33; p=0.0719) and MQ3+MQ4 (F(1, 76) =3.11;
p=0.0821).
2. Interaction Within Polishing Direction T-C

Results of two-way ANOVA (MQ1) and two-way ANOVA based on
ranked data (MQ3+MQ4) revealed a statistically significant main effect of
the type of resin-based composite for MQ1 (F(1,76)=6.94; p=0.0102) and
MQ3+MQ4 (F(1,76)=5.50; p=0.0216). The post-hoc Tukeys HSD test
101
indicated that the mean MQ1 observed in FSU was significantly greater than
that in PT4 (mean MQ1: 83.8912.35 vs. 76.5012.75, respectively). The posthoc Bonferroni test indicated that the mean MQ3+MQ4 observed with PT4
was significantly greater than that with FSU (mean MQ3+MQ4: 7.9310.57
vs. 4.347.36, respectively). Two-way ANOVA based on ranked data showed
no significant main effect for the type of resin-based composite for MQ2 (F(1,
76)=2.25, p=0.1373).
These analyses also proved the main effect for the polishing technique
to be non-significant for MQ1 (F(1,76)=0.96, p=0.3309), MQ2 (F(1,76)=2.43;
p=0.1229) and MQ3+MQ4 (F(1,76)=1.00, p=0.3196).
The
interaction
between
resin-based
composite
and
polishing
technique also proved to be non-significant for MQ1 (F (1,76)=1.32;

p=0.2543), MQ2 (F(1, 76) =0.35; p=0.5537) and MQ3+MQ4 (F (1,76) =0.95;
p=0.3320).
Interaction between Polishing Technique

and Polishing Direction for Each
Resin-Based Composite
When taking into account the correlated data obtained from the same
specimens, correlations among measurements made on the same specimen
can be modeled using the random effects in Mixed Model ANOVA (i.e. to
allow correlation between two measurements of polishing direction obtained
102
from the same specimen) to evaluate the effects of polishing techniques and
polishing directions, including an interaction between the two factors.
Since the data were not normally distributed, random effects in Mixed
Model ANOVA based on ranked data were performed. Comparisons were
made for FSU and PT4.
1- Interaction Within FSU

Results revealed a statistically significant main effect for the polishing
technique for MQ1 (F(1, 38)=11.65; p=0.0015) and MQ2 (F(1, 38)=19.68;
p<0.0001). No significant difference was found for MQ3+MQ4 (F(1,38)=1.00,
p=0.3226)). The post-hoc Bonferroni test indicated that the mean MQ1
observed with R was significantly greater than that observed with SL (mean
MQ1: 92.3110.59 vs. 86.1412.53, respectively), and that the mean MQ2
observed with SL was significantly greater than that observed with R (mean
MQ2: 11.2512.03 vs. 5.467.55, respectively).
Results also revealed a statistically main effect for the polishing
direction for MQ1 (F(1, 38)=28.72; p<0.0001), MQ2 (F(1, 38)=14.33; p=0.0005)
and MQ3+MQ4 (F(1, 38)=19.68; p<0.0001). The post-hoc Bonferroni test
indicated that the mean MQ1 for polishing direction C-T was greater than for
polishing
direction
T-C
(mean
MQ1:
94.568.82
vs.
83.8912.35,
respectively), the mean MQ2 for polishing direction T-C was greater than for
polishing direction C-T (mean MQ2: 11.7711.17 vs. 4.938.37, respectively),
and that the mean MQ3+MQ4 observed for polishing direction T-C was
103
greater than for polishing direction C-T (MQ3+MQ4: 4.347.36 vs. 0.501.95,
respectively).
The interaction between polishing technique and polishing direction
was
non-significant
for
MQ1
(F (1, 38) =0.51; p=0.4783),
MQ2
(F (1, 38) =0.82; p=0.3696) and MQ3+MQ4 (F (1,38) =0.91; p=0.3468).
2- Interaction Within PT4

When restoring with PT4, results revealed a statistically significant
main effect for polishing direction for MQ1 (F(1, 38)=127.14; p<0.0001), MQ2
(F(1, 38)=70.48; p<0.0001) and MQ3+MQ4 (F(1, 38)=74.35; p<0.0001). Posthoc Bonferroni tests indicated that the mean MQ1 observed for polishing
direction C-T was greater than for polishing direction T-C (mean MQ1:
97.743.93 vs. 76.5012.75, respectively), the mean MQ2 observed for
polishing direction T-C was greater than for polishing direction C-T (mean
MQ2: 15.5711.80 vs. 2.223.79, respectively) and that the mean MQ3+MQ4
observed for polishing direction T-C was greater than for polishing direction
C-T (mean MQ3+MQ4: 7.9310.57 vs. 0.040.18, respectively).
The main effect for polishing technique proved to be non-significant for
MQ1
(F(1,38)=0.19,
p=0.6693),
MQ2
(F(1,38)=0.88,
p=0.3540))
and
MQ3+MQ4 (F(1,38)=2.60, p=0.1149).

The interaction between the polishing technique and the polishing
direction also proved to be non-significant for MQ1 (F (1,38) =0.33; p=0.5698),
MQ2 (F(1,38) =0.04; p=0.8477) and MQ3+MQ4 (F (1,38) =2.44; p=0.1262).
104
Interaction between Resin-Based Composite
and Polishing Direction for Each
Polishing Technique
When taking into account the correlated data obtained from the same
specimens, correlations among measurements made on the same specimen
allow correlation between two measurements of polishing direction obtained
from the same specimen) to evaluate the effects of the type of resin-based
composite and the polishing direction, including an interaction between the
two factors.
1- Interaction Within Polishing Technique SL

Model ANOVA based on ranked data was performed. This analysis revealed
a statistically significant interaction between the type of resin-based
composite and the polishing direction for MQ1 (F(1, 38)=6.62; p=0.0141), but
proved to be non-significant for MQ2 (F (1,38) =4.03; p=0.0519) and
MQ3+MQ4 (F (1,38) =2.55; p=0.1189). Subsequent analyses for detecting
simple effects for the type of resin-based composite and for the polishing
direction in MQ1 were conducted.
Based on Wilcoxon rank-sum test, there was a simple effect for the
type of resin-based composite at polishing direction C-T (p=0.0160) and it
indicated that the mean MQ1 for PT4 was significantly greater than that for
105
FSU (mean MQ1: 96.965.17 vs. 91.3811.02, respectively). However, based
on two-sample t-test, the simple effect for the type of resin-based composite
at the polishing direction T-C was non-significant (i.e. there was no
difference between the two resin-based composites) (p=0.2782).
Based on a paired-sample t-test, there was a simple effect for the type
of polishing direction within FSU (p=0.0055) and within PT4 (p<0.0001).
Data showed that the mean MQ1 for polishing direction C-T was
significantly greater than for polishing direction T-C within FSU (mean
MQ1: 91.3811.02 vs. 80.9111.97, respectively) and within PT4 (mean MQ1:
96.965.17 vs. 76.7411.98, respectively).
Results of Mixed Model ANOVA based on ranked data also revealed a
statistically significant main effect for polishing direction for MQ2
(F(1, 38)=24.89; p<0.0001) and MQ3+MQ4 (F(1, 38)=25.83; p<0.0001). The
post-hoc Bonferroni test indicated that the mean MQ2 observed for polishing
direction T-C was significantly greater than for polishing direction C-T
(mean MQ2: 15.9312.54 vs. 5.308.49, respectively) and the mean
MQ3+MQ4 observed for polishing direction T-C was significantly greater
than for polishing direction C-T (mean MQ3+MQ4: 5.248.15 vs. 0.531.95,
respectively).
The main effect for the type of resin-based composite proved to be nonsignificant
for
MQ2
(F(1,38)=0.01, p=0.9345)).
(F(1,38)=1.03,
p=0.3176)
and
MQ3+MQ4
106
2- Interaction Within Polishing Technique R
Model ANOVA based on ranked data was performed. This analysis revealed
a statistically significant main effect of the type of resin-based composite for
MQ1(F(1, 38)=4.47; p=0.0410) while for MQ2 it was proven to be nonsignificant (F(1,38)=1.08, p=0.3051). The post-hoc Bonferroni test indicated
that the mean MQ1 observed for FSU was significantly greater than that
observed for PT4 (mean MQ1: 92.3110.59 vs. 87.3914.88, respectively).
Results from this analysis also showed a statistically significant main
effect of polishing direction for MQ1 (F(1, 38)=72.96; p<0.0001) and MQ2
(F(1, 38)=46.53; p<0.0001). The post-hoc Bonferroni test indicated that the
mean MQ1 for polishing direction C-T was greater than that for polishing
direction T-C (mean MQ1: 98.133.20 vs. 81.5713.97, respectively) and the
mean MQ2 observed for polishing direction T-C was greater than that for
polishing direction C-T (mean MQ2: 11.4110.19 vs. 1.853.17, respectively).
The interaction between the type of resin-based composite and
polishing direction was non-significant for MQ1 (F (1, 38) =3.23; p=0.0801)
and MQ2 (F (1,38) =1.16; p=0.2879). However, a statistically significant
interaction between the type of resin-based composite and the polishing
direction was found for MQ3+MQ4 (F(1, 38)=4.79; p=0.0348). Subsequent
analyses for detecting simple effects for resin-based composite and for
polishing direction were conducted.
107
The analysis using Wilcoxon rank-sum test demonstrated that there
was a simple effect for the type of resin-based composite at polishing
direction T-C (p=0.0381) and it indicated that the mean MQ3+MQ4 for PT4
was significantly greater than for FSU (mean MQ3+MQ4: 9.5712.02 vs.
4.487.50, respectively). The simple effect of the type of resin-based
composite at the polishing direction C-T was non-significant (i.e. there was
no difference between the two resin-based composites) (p=0.3421) (Table 6).
The analysis using a paired-sample t-test revealed that the simple
effect of the type of polishing directions within FSU was significant
(p=0.0151). The data showed that the mean MQ3+MQ4 for polishing
direction T-C was significantly greater than that for polishing direction C-T
(mean MQ3+MQ4: 4.487.50 vs.0.000.00, respectively). The simple effect of
the type of polishing directions within PT4 also proved to be significant
(p=0.0022). Data showed that the mean MQ3+MQ4 for polishing direction TC was significantly greater than that for polishing direction C-T (mean
MQ3+MQ4: 9.5712.02 vs.0.040.16, respectively) (Table 6).
Interaction between Resin-Based Composites,

Polishing Technique and Polishing Direction
When taking into account the correlated data obtained from a same
specimen, correlations among measurements made on the same specimen
allow correlation between two measurements of polishing directions obtained
108
from the same specimen) to evaluate the effects of the type of resin-based
composite, the polishing technique and the polishing direction, including
interactions between the three factors.
For Variable Continuous Margin (MQ1)

This analysis revealed that there were statistically significant main
effects for the polishing technique (p=0.0399) and for the polishing direction
(<0.0001). The results demonstrated significant differences between the two
polishing techniques and the two polishing directions. The interaction
between the type of resin-based composite and the polishing direction proved
to be significant (p=0.0011) and indicated that the polishing directions had
significant differences between the two types of resin-based composite.
However, the results showed a non-significant effect for the type of
resin-based composite (p=0.1931), and non-significant interactions between
the type of resin-based composite and the polishing technique (p=0.0831),
between the polishing technique and the polishing direction (p=0.6970), and
between the type of resin-based composite, the polishing technique and the
polishing direction (p=0.7946). Table 13 displays the results of the random
effects in Mixed Model ANOVA.
109
Table 13. Summary of the Full Model Used in the Random Effects in Mixed
Model ANOVA for MQ1
Source
df
Type III
F-value
p-value
Resin-based composite
1.72
0.1931
Polishing technique
4.37
0.0399*
Polishing direction
105.40
<0.0001*
Resin-based composite by polishing

direction
11.55
0.0011*

technique
3.08
0.0831
Polishing technique by polishing

technique
0.15
0.6970

technique by polishing direction
0.07
0.7946
For Variable Marginal Irregularity (MQ2)

Mixed Model ANOVA revealed statistically significant main effects for
the polishing technique (p=0.0101) and for the polishing direction (<0.0001).
The results demonstrated significant differences between the two polishing
techniques and between the two polishing directions. Significant interaction
between the type of resin-based composite and the polishing direction
(p=0.0217) were found and indicated that the polishing directions had
significant differences between the two types of resin-based composite.
However, the results showed non-significant effects for the type of
resin-based composite (p=0.7202), non-significant interactions between the
type of resin-based composite and the polishing technique (p=0.2358),
110
between the polishing technique and the polishing direction (p=0.7002), and
between the type of resin-based composite, the polishing technique and the
polishing direction (p=0.9414). Table 14 displays the results of the Random
effects in Mixed Model ANOVA.
Model ANOVA for MQ2
Source
df
Type III
F-value
p-value
0.13
0.7202
Polishing technique
6.95
0.0101*
Polishing direction
52.76
<0.0001*

direction
5.50
0.0217*

technique
1.43
0.2358

technique
0.15
0.7002

0.01
0.9414
For Variable Gaps (MQ3+MQ4)

Mixed Model ANOVA revealed that there was statistically significant
main effects for the polishing direction (p<0.0001). The results demonstrated
a significant difference in MQ3+MQ4 between the two polishing directions.
111
However, the results showed non-significant main effects for the type
of resin-based composite (p=0.1365) and for the polishing technique
(p=0.5401), non-significant interactions between the type of resin-based
composite and the polishing direction (p=0.0535), between the type of resinbased composite and the polishing technique (p=0.3347), between the
polishing technique and the polishing direction (p=0.2702), and between the
type of resin-based composite, the polishing technique and the polishing
direction (p=0.6240). Table 15 displays the results of the Random effects in
Mixed Model ANOVA.
Model ANOVA for MQ3+MQ4
Source
df
Type III
F-value
p-value
2.26
0.1365
Polishing technique
0.38
0.5401
Polishing direction
32.31
<0.0001*

direction
3.85
0.0535

technique
0.94
0.3347

technique
1.23
0.2702

0.24
0.6240
112
Statements on Research Hypotheses
Table 16. Statements on Main Hypotheses
Main Hypothesis
Conditions
Conclusion
Ho (1): there is no difference in marginal adaptation

between polishing direction C-T and polishing
direction T-C under different conditions
Rejected
Ho (2): there is no difference in marginal adaptation between

polishing technique SL and polishing technique R
under different conditions.
Ho (3): there is no difference in marginal adaptation between
resin-based composite FSU and resin-based composite
PT4 under different conditions.
Rejected
Rejected
Table 17. Statements on Secondary Hypotheses

Secondary Hypothesis
Conditions
Conclusion
Ho (1.2): Resin-based composite has no influence on how

polishing technique affects marginal adaptation
within each polishing direction.
Rejected
Ho (2.2): Polishing technique has no influence on how

polishing direction affects marginal adaptation
within each resin-based composite.
Rejected
Ho (3.2): Resin-based composite has no influence on how

polishing direction affects marginal adaptation
within each polishing technique.
Rejected
Ho (4.2): There is no interaction among the three factors:

resin-based composite, polishing technique and
polishing direction.
Rejected
113
CHAPTER V
DISCUSSION
Resin-based composite materials are widely used for anterior direct

esthetic restorations. Although finishing and polishing are important to
prevent plaque retention, marginal staining, gingival irritation and recurrent
caries, these procedures can also create marginal defects and gaps that can
be detrimental to longevity and esthetics. The most common reason for resinbased composite failure reported in the literature is recurrent caries (Mjor
and Qvist 1997; Forss and Widstrom, 2004; Bernardo et al., 2007). Therefore,
maintaining the integrity of the margins by applying a proper finishing and
polishing protocol is an important goal to pursue.
This in vitro study evaluated the effect of finishing and polishing
direction (from C-T and from T-C) on marginal adaptation of resin-based
composite restorations using two finishing and polishing techniques and two
resin-based composite materials.
Previous studies evaluating the effect of finishing and polishing
procedures on microleakage or marginal adaptation did not investigate
whether the finishing/polishing direction from resin-based composite to tooth
structure or from tooth structure to resin-based composite made any
difference in the quality of the margins. Some opinion leaders advocate in
lectures and conferences to finish and polish from resin-based composite to
114
tooth structure, but there is no scientific evidence in the dental literature to
support this clinical recommendation.
Marginal Adaptation vs. Microleakage

Marginal adaptation and microleakage are two methods to measure or
evaluate the marginal seal of restorations. While microleakage allows
observation of the depth of marginal defects, marginal adaptation allows
measurement of their width and identification of other characteristics such
as marginal irregularities. Also, microleakage studies are based on
qualitative assessment of defined criteria without a quantitative component.
Therefore, they rely more on subjectivity. In addition, microleakage studies
evaluate only a small portion of the margin compared to marginal adaptation
studies where the complete margin can be assessed.
In this present study, marginal adaptation was chosen instead of
microleakage because of the possibility to quantify a margin length of 2 mm
according to specific criteria: continuous margin, marginal irregularity,
hairline crack and severe gap. Moreover, because this method of evaluation is
non-destructive, it could be repeated in vivo and correlated with this in vitro
study.
Since many studies evaluating the marginal integrity of resin-based
composite restorations after finishing and polishing procedures are based on
microleakage, comparisons of results with this study are difficult and
similarities or disparities should be interpreted with caution.
115
Quantitative Margin Analysis
Quantitative margin analysis, introduced in 1984 by Porte et al. (1984)
and further developed in 1989 by Roulet et al. (1989) and Blunck et al. (1989)
was chosen in this study for quantitative statistical comparisons and to
reduce subjectivity.
This method is commonly used in marginal adaptation studies and can
be defined with as few criteria as two (pass or fail) (Frankenberger et al.,
2008) or as many as seven (Sabatini, 2007; Sabatini et al., 2010). However,
when many criteria are involved, categories are often collapsed to simplify
statistical analysis and interpretation of the results. As an example, Sabatini
et al. (Sabatini, 2007; Sabatini et al., 2010) defined seven different marginal
quality criteria ranging from a perfect margin to a gap width of more than 5
microns. In their presentation and analysis of results, the authors further
reduced the number of categories into four and then furthermore into two
categories based on whether there was a gap or not. Blunck et al. (Blunck
and Roulet, 1989; Blunck and Zaslansky, 2011) performed a quantitative
margin analysis based on four marginal quality criteria. The number of
criteria is a very important factor. An analysis based on pass or fail criteria
can be misleading because it does not differentiate between marginal defects
that can have a very different clinical impact on the longevity of the
restoration, such as a marginal irregularity versus a marginal gap. A large
number of criteria can lead to reliability issues and may increase the
subjectivity component during the margin analysis. Roulet et al. (1989)
116
reported a difference in the accuracy of the method when using five or four
criteria. The authors reported that the difference between two measurements
taken in a four-week interval by a same operator when using five criteria
was 3% 2.6% with the largest difference being 9% while it was 1.9% 0.9%
with the largest difference being 3.4% when using four criteria. In our study,
it was decided to limit the number of criteria to four as described by Blunck
et al. (Blunck and Roulet, 1989; Blunck and Zaslansky, 2011) in order to be
able to characterize the marginal integrity specifically, yet objectively. The
four categories were further collapsed into three categories for statistical
purposes and to simplify the interpretation of the results obtained.
In our study, intra-observer reliability of the measurements was tested
for all the specimens included in the pilot study. Eighty-four margins were
measured blindly twice in a two-week interval. A paired-sample t-test did not
reveal significant differences between the two measurements made on the
same margin by a single observer and intraclass correlation coefficient
indicated strong agreement for all four marginal quality criteria. This is in
partial agreement with the Sabatini et al., (Sabatini, 2007) study which
reported mostly high and moderately high level of agreement between two
measurements. However, the authors also reported poor level of agreement
for two of their margin criteria. This can be explained by the fact that a large
number of criteria were used in their study and could have led to reliability
issues.
117
Replicas Technique
Positive replicas were used in this study for observation of the margins
with FeSEM. Tayor et al. (1993) reported that this technique is frequently
used and accepted in marginal adaptation studies involving SEM. It has also
been stated that using replicas is a more precise and reliable technique than
imaging dental tissues since the latter involves specimen dehydration which
may cause dental tissues to shrink leading to a gap that appears wider than
it actually is (Davila et al., 1989). Also, cracks caused by the high vacuum
may create artifacts or appear as a gap at the margin.
One disadvantage of this replica technique, as experienced in our
study, is the incorporation of bubbles in the replicas. Bubbles situated on the
margins were considered as artifact and subtracted from the total length of
the studied margin. This reduced the total length of the studied margin and
may have hidden possible gaps or marginal irregularities. Precautions were
taken to reduce bubbles. First, impressions rested for 24 hours before being
poured with epoxy resin to allow gas to escape from the impression material
during setting. Additionally, freshly mixed epoxy resin was placed in a
vacuum for 10 minutes to remove the bubbles resulting from the mixture.
Even with these precautions, bubbles could not be completely avoided when
looking at 200X magnification. Maybe a longer period of time in a more
powerful vacuum would have helped avoiding bubbles. However, bubbles
situated outside the margins were helpful when imaging with SEM. They
118
helped making sure that the images taken were precisely overlapping each
other and facilitated merging of images with Photoshop software.
Control of Psychomotor Skills Variability

To control for the variability of hand skills throughout the study, the
operator rehearsed prior to the study to obtain a standardized speed and
pressure while finishing and polishing the resin-based composite. Moreover,
a random order list of specimens groups was followed to restore, finish and
polish one specimen in each group before moving forward to the next ones.
Variables Affecting Gap Formation

During the restorative process, gap formation can be caused by
polymerization shrinkage which is affected by the cavity size and shape
(Taylor and Lynch, 1993). The C-factor, estimating the amount of stress
generated at the tooth-restoration interface, is the ratio of bonded to
unbonded surfaces (Feilzer et al., 1987). As the C-factor increases, more
stress is developed in the material during its polymerization, increasing the
risk of gap formation. In the present study, cavity preparation was done in a
triangular shape with a depth of only 0.87 mm. The resin-based composite
was bonded to no more than two surfaces. Therefore, the polymerization
shrinkage was minimized.
Conversely, hygroscopic expansion occurs when resin-based composite
restorations are placed in water or saliva. Hygroscopic expansion causes
119
swelling of resin-based composite and may improve the marginal seal (Yap et
al., 2003). In our study, in order to minimize the effect of hygroscopic
expansion, the specimens were stored in artificial saliva no longer than 4
hours before taking the impressions to make replicas. Moreover, because of
the way the specimens were finished and polished in a random order, each
specimen had equal chances to be in artificial saliva for a prolonged period of
time of up to 4 hours.
Cavity Preparation and Restoration

Prior to the beginning of the study, the methodology was tested with
an infinite bevel instead of a 30 bevel because, in esthetic zones, a longer
bevel allows a more harmonious and esthetic blend between the resin-based
composite and the tooth structure (Summitt, 2006). However, this design did
not allow the detection of gap and therefore a more defined finish line
obtained with a 30 bevel was preferred for this study.
Some microleakage studies (Lopes et al., 2002; Yalcin et al., 2006;
Cenci et al., 2008) have shown differences between dentin and enamel
margins after finishing and polishing procedures. Therefore, while preparing
the specimens for this experimentation, careful attention was made to avoid
exposing dentin while flattening the enamel in order to ensure that all
margins would be located in enamel only.
The main goal of the study was to assess the effect of polishing
direction. However, finishing and polishing are closely related and one
120
cannot be accomplished without the other. When restoring the specimens,
the resin-based composite was applied in one increment and carefully
adapted to minimize excess in such a way that finishing procedures were
reduced to a minimum.
The margins of the resin-based composite restorations on the enamel
surface were as straight as possible. Sometimes, under a magnification of
200X, the margins appeared slightly wavy. For better consistency, the
analysis and measurement of the margins needed to be done in a straight
line. Therefore a straight line was drawn as a reference next to the margins
and all measurements such as margin length and marginal defects were done
linearly and parallel to that reference line.
Effect of Finishing and Polishing Direction

The results of our study showed that in most instances, more
continuous margins, less marginal irregularities and less gaps were found
when the finishing and polishing procedures were performed from C-T rather
than from T-C. However, one group (FSU/SL) did not show a significant
difference in marginal gaps, but the p-value was 0.0537 and can be
suggestive that more gaps were present in T-C direction in that group as
well.
One explanation might be that while finishing and polishing from C-T
the resin-based composite may be pushed against the enamel margin,
protecting the marginal seal from breakage compared to finishing and
121
polishing from T-C where the resin-based composite can be pulled away from
the margin, challenging the adhesive bond at the tooth-restoration interface
and
increasing
the
susceptibility
to
gap
formation
and
marginal
irregularities.
Moreover, it has been reported that only 75% of the material is cured
after 10 minutes and thus, immediate finishing and polishing may cause
plastic deformation (Powers et al., 2006). Therefore, the incomplete
polymerization of the material and the heat generated during finishing and
polishing procedures may have caused an increased level of plasticity which
could have been an advantage for C-T direction by preserving the marginal
seal and a disadvantage for T-C direction by leading to marginal breakdown.
Another possible reason could be that fragile enamel rods at the border
of the finish line are more susceptible to breakdown and may be wrenched
away from the enamel bulk with a T-C direction while preserved with a C-T
direction.
Different studies have demonstrated that finishing procedures may
cause gap formation at the tooth-restoration interface (Yu et al., 1990;
Brackett et al., 1997; Yap et al., 1998; Lopes et al., 2002; Schmidlin and
Gohring, 2004; Yalcin et al., 2006; Cenci et al., 2008; Maresca et al., 2010)
which is in agreement with the present study where gaps and marginal
irregularities were observed.
In a study conducted by Maresca et al (2010), there was no significant
difference between parallel or perpendicular orientation of finishing
122
procedures. However, the perpendicular direction was performed using the
same mesio-distal orientation without distinction between finishing from the
C-T or from T-C according to the direction of the bur grits or blades.
Effect of Finishing and Polishing Technique

While there was mostly no difference between SL and R under the
conditions tested, differences were observed in one group for all marginal
quality criteria. Indeed, a difference between SL and R was found when FSU
was finished and polished from C-T. The results revealed more continuous
margins, less marginal irregularities and less gaps when rubber polishers
were used instead of Sof-Lex discs.
Regarding marginal irregularities, a possible explanation could be that
abrasive particles on Sof-Lex discs are strongly bonded to the discs and are
not easily dislodged while finishing or polishing, creating scratches in the
resin-based composite that are not reduced and smoothed down by the
subsequent discs of lower abrasiveness. This finding was also reported by
Van Noort et al. (1984) who mentioned that abrasive particles in Sof-Lex
discs, because they do not follow the path of least resistance as it would be
the case in polishing pastes, dislodge filler particles from the material
creating scratches or deep notches practically impossible to remove with
subsequent discs of smaller abrasive particles. The authors also mentioned
that this phenomenon is more common with resin-based composite
123
containing small particle size, which may explain why more marginal
irregularities were found with FSU than with PT4 in the present study.
However, differences in marginal defects between the two finishing
and polishing technique were found only for the C-T direction. This can be
explained by the fact that in the T-C direction, the abrasive particles are in
contact with enamel first at a certain speed which may have protected the
surface of the resin-based composite immediately adjacent to the toothrestoration interface. After a certain distance, on the surface of the resinbased composite, which is softer than enamel, abrasive particles may have
created deep scratches at a lower level than the adjacent enamel structure
leaving irregularities difficult to remove with subsequent discs of finer grits.
Regarding marginal gaps, one possible reason why more gaps were
found in the Sof-Lex group compared to rubber polishers group may be
related with the heat generated while finishing and polishing. In the present
study, no water was used with Sof-Lex discs while water was used with one
of the polishing steps in the rubber polishers group. The absence of water
may have caused a temperature elevation producing a breakdown at the
tooth-restoration interface. This is in agreement with previous studies in
which authors have reported the negative effect of heat on the adhesive bond
at the tooth-restoration interface (Baratieri L. N. et al., 1998; Yap et al.,
1998; Powers et al., 2006). Yu et al. (1990) have also mentioned that while
finishing with discs without water, the amount of heat may create gaps
because of the difference in coefficient of thermal expansion between the
124
resin-based composite and the tooth structure. However, in that study, the
authors did not find microleakage at the enamel margins while considerable
microleakage occurred in dentin margins finished with dry discs. This latter
finding by Yu et al. is not in accordance with the results of the present study
where marginal gaps were found in enamel margins.
Effect of Type of Resin-based Composite

While there was mostly no difference between FSU and PT4 under the
conditions tested, a few differences were observed. PT4 showed more gaps
and less continuous margins than FSU when polished from T-C with rubber
polishers. Also, PT4 showed more continuous margins and less marginal
irregularities when polished from C-T with SL. PT4 seemed to be more
affected by the polishing direction and therefore, the effect of finishing and
polishing directions appears to be material dependent and cannot be
generalized to all resin-based composites.
These findings are not in accordance with those from previous studies
comparing nanofilled and microhybrid resin-based composites. Yalcin et al.
(2006) found no significant difference in microleakage of enamel margin with
various types of materials (nanofilled, nanohybrid and microhybrid) and
polishing systems (Super-Snap disks, Astropol/Astrobrush polishing system).
Dentin margins, however, showed significant differences with more leakage
occurring in the microhybrids followed by the nanohybrids and then the
nanofills.
125
According to Asmussen et al. (1998), it is difficult to explain
differences in mechanical and physical properties between resin-based
composites because they vary in many characteristics such as their matrix,
their filler particles (content, size and morphology) and their particle-matrix
coupling. FSU and PT4 differ in the constituents of their resin matrix and in
their particles characteristics. While the filler content of FSU and PT4 is
comparable and respectively 72.5% and 76% by weight, the particle size and
morphology are different. FSU contains spherical particles of 20 nm (silica),
4-11 nm (zirconia) and 0.6-1.0 micron (clusters). PT4 contains irregular
particles of glass and silicone dioxide with an average particle size of 0.4
micron.
Differences between the Eight Experimental Groups

When comparing the marginal adaptation between the eight
experimental groups, PT4 finished and polished with a diamond bur followed
by a Sof-Lex disc and rubber polishers from C-T showed one of the best result
of all groups for both continuous margin and marginal gaps criteria.
Interestingly, this same resin-based composite, finished and polished with
the same technique but from T-C direction was among the worse results in
the study for both continuous margins and marginal gaps. This suggests that
the marginal adaptation of PT4 is highly dependent on the finishing and
polishing direction. PT4 seems to be more susceptible to be pulled away from
the margins with a T-C direction while keeping a good seal with a C-T
126
direction. This is more likely to be due to the filler and matrix characteristics
of PT4 that directly influence mechanical and physical properties when
compared to FSU that seems to present a more consistent behavior.
Interactions
No interactions were found between the type of resin-based composite
and the polishing technique for all marginal quality criteria. However, the
type of resin-based composite was found to have a main effect, indicating
that more continuous margins and less marginal gaps were found with FSU
than with PT4 under the polishing direction T-C. This finding is most likely
to be related to the resin-based composite formulation and its mechanical
and physical properties. The polishing technique also proved to have a main
effect where more continuous margins and less marginal irregularities were
found with the polishing technique R. This is in agreement with the analysis
explained earlier where polishing technique SL led to more marginal
irregularities.
No interactions were found between the polishing techniques and the
polishing directions. However, the polishing techniques were found to have a
main effect under FSU. More continuous margins and less marginal
irregularities were found when using the polishing technique R with FSU.
This finding is also in agreement with the previous analysis. The polishing
direction was also found to have a significant main effect for all marginal
quality criteria. Systematically, the polishing direction C-T led to more
127
continuous margins, less marginal irregularities and less marginal gaps than
the polishing direction T-C which is also in agreement with the findings of
the analysis explained earlier.
Interactions were found between the resin-based composite and the
polishing directions within the polishing technique SL for MQ1 and within
the polishing technique R for MQ3+MQ4. This means that the relationship
between the polishing direction and MQ1 is different for the two resin-based
composites when using the polishing technique SL. Within the polishing
technique SL, the simple effect for the type of resin-based composite with a
direction C-T indicated that PT4 led to more continuous margins than FSU.
This is in agreement with previous analyses. The relationship between the
polishing direction and MQ3+MQ4 is also different for the two resin-based
composites when using the polishing technique R. Indeed, within the
polishing technique R, for polishing direction T-C, PT4 showed more
marginal gaps than FSU which is also in agreement with previous analyses.
The relationship between the resin-based composite and MQ1 showed a
difference for the two polishing directions within the polishing technique SL.
A simple effect of polishing direction indicated that polishing direction C-T
achieved more continuous margins within polishing technique SL and less
marginal gaps within polishing technique R for both FSU and PT4. Polishing
direction also showed a main effect under polishing technique SL and R,
indicating that the direction T-C led to more marginal irregularities and gaps
than the polishing direction C-T for SL and R as well as more continuous
128
margins in the polishing technique R than SL. The type of resin-based
composite also showed a main effect within the polishing technique R,
indicating that more continuous margins were observed in FSU than PT4.
All of these findings are in agreement with previous analyses.
While no interactions were found among the three independent
variables, a main effect for the polishing technique and polishing directions
were detected. Significant differences were found in continuous margins and
marginal irregularities between the two polishing techniques and the two
polishing directions. Only the polishing direction had a main effect regarding
marginal gaps. An interaction was also found only for continuous margins
between the resin-based composite and the polishing direction indicating
that polishing directions were different between the two types of resin-based
composite. All of these findings are in agreement with the previous analysis.
Control of Possible Bias

Possible bias may result from the preparation design, the finishing
and polishing procedures, the viewing angle during FeSEM observations and
the measurement methodology.
The preparation was standardized with a depth of 0.87 mm and a
width of 3 mm in order to obtain 30 bevels. All steps were done by the same
operator who rehearsed prior to the finishing and polishing procedures in
order to maintain a constant pressure and speed. However, no device was
129
used to monitor the pressure, such as the custom-made one used in Maresca
et al. study (2010) to ensure a standardized pressure of 0.5N.
When imaging with SEM, specimens should be viewed in a parallel
plane. A small angulation change may result in a marginal gap appearing
smaller than it actually is (Taylor and Lynch, 1993). In the present study,
replicas were carefully mounted with margins oriented parallel to the stub
base. Therefore, the margins observed with FeSEM were perpendicular to
the electron beam and should not have undergone deformation due to the
angulation.
All measurements were taken blindly in order to limit observers bias.
For reliability purposes, measurements were taken twice in a two-week
interval for all specimens included in the pilot study. With 0.05 level of
statistical significance, the results showed very strong agreement with an
intraclass correlation coefficient of 0.99 for MQ1, 0.99 for MQ2, 0.97 for MQ3,
and 0.99 for MQ4.
Strengths of the Study

The present study performed a power analysis using 6 specimens per
group to determine the sample size requirement of 20 specimens per group
(0.05 level of significance and 80% power).
Comparable studies evaluating marginal adaptation of resin-based
composite with SEM and using similar marginal quality criteria used a
smaller sample size. Sabatini et al. (Sabatini, 2007; Sabatini et al., 2010),
130
Blunck et al. (2011) and Frankenberger et al. (2008) all used 8 specimens per
group. Maresca et al. (2010) used 5 specimens per group. The present study
had 20 specimens per experimental group.
Moreover, cavity preparations in most studies assessing the effect of
finishing and polishing procedures on marginal adaptation of resin-based
composite have a 90cavosurface angle. For anterior teeth where esthetic is
important, a bevel is usually made clinically to allow a smooth shade
transition between the resin-based composite and the tooth structure (Albers,
2002). Therefore, this study is closer to what would normally be done
clinically for anterior restorations.
Limitations of the Study

This in vitro study included some manipulations that are not normally
performed in a clinical situation. For instance, the enamel was flattened to
allow a better definition of the margins with FeSEM. It has been shown that
the outer surface of enamel may contain aprismatic enamel (Ripa et al., 1966;
Whittaker, 1982) and is often hypermineralized with high fluoride content
(Kanemura et al., 1999; Albers, 2002). This layer was removed by flattening
the enamel. However, Kanemura et al. (1999) found no significant difference
in tensile bond strength between ground and intact enamel when using
phosphoric acid. Also, the two 30 bevels obtained as a result of the
standardized triangular shape preparation is not a clinical common
preparation. Moreover, the quality of the polished surface was observed with
131
20X magnification which is not typically done clinically unless the dentist
works with a microscope. Furthermore, the design of the standardized
preparation facilitated the access to the margins for finishing and polishing.
In clinical situations, margins are frequently difficult to access gingivally and
interproximally and may not be polishable from all directions.
In the present study, only margins in enamel were assessed. It is
believed that margins in dentin would likely lead to different results and
therefore, the results cannot be generalized to all clinical situations. Also,
there are several resin-based composites and adhesive systems available on
the market. For this study, only Filtek Supreme Ultra and Point 4 as resinbased composites and only OptiBond FL as adhesive were evaluated. Results
should be interpreted with caution and may not apply to other materials.
Clinical Significance
A statistically significant difference was found in this study regarding
the direction of finishing and polishing procedures revealing that the C-T
direction leads to more continuous margins with less marginal irregularities
and less marginal gaps than T-C. Marginal adaptation was evaluated in an
in vitro setting where the margins were easily accessible for finishing and
polishing for both directions tested. Finishing and polishing from the resinbased composite to the tooth structure is not always possible clinically and
cannot be done in all circumstances. As mentioned earlier, margins are often
difficult to access interproximally or close to the gingival margins. The
132
results obtained from this in vitro study suggest that finishing and polishing
from the resin-based composite to the tooth structure should be clinically
achieved whenever possible.
Regarding the type of resin-based composite, Point 4 seems to be more
susceptible to marginal gaps with a polishing direction T-C compared to
FSU. However, PT4 has also better marginal adaptation with a C-T direction
than FSU. This suggests that when margins are accessible for finishing and
polishing with a C-T direction, Point 4 is a very good material to use whereas
FSU may be a better alternative when the margins are difficult to access
from C-T direction.
Regarding the polishing technique, there was overall no major
difference between polishing techniques SL and R. However, the polishing
technique involving rubber polishing cups may have an advantage over SofLex discs since Sof-Lex discs have led to more marginal irregularities and
gaps with a C-T direction in one group.
Regarding the clinical relevance of marginal gaps, the dental
literature is quite controversial about the association between gap sizes and
recurrent caries. While some in vitro studies have demonstrated a positive
relationship between gap formation and recurrent caries (Totiam et al., 2007)
other in situ studies considering more factors such as fluoride and patients
risk factors have concluded that a gap does not influence the formation of
recurrent caries (Cenci et al., 2008; Cenci et al., 2009). Also, the dimension of
a gap that could potentially cause recurrent caries remains unknown.
133
Although it has been reported that a gap as small as 0.5-1.0 m wide at the
tooth-restoration interface would allow bacterial infiltration and that smaller
gaps can allow toxin and bacterial byproduct infiltration (Taylor and Lynch,
1992), no clear gap dimensions are stated. Totiam et al. (2007) assessed the
association between different gap sizes and recurrent caries and reported
that none of the gap sizes included in their study completely avoided
recurrent caries. Some studies have reported no association between gaps
and recurrent caries on the basis of fluoride exposure and patients oral
hygiene (Sarrett 2007; Cenci et al., 2008; Cenci et al., 2009). However, even if
a gap does not necessarily lead to recurrent caries, it still can cause marginal
staining which is often considered as a failure for esthetic anterior
restorations.
Finally, many different factors influence the marginal seal of resinbased composite restorations during their lifetime. It has been shown that
the seal and marginal adaptation of a restoration undergo degradation with
time because of masticatory forces, thermal changes and hydrolysis (GarciaGodoy et al., 2010). Perfect marginal adaptation immediately after placement
and finishing is likely to increase longevity of restorations, but it is not
sufficient to prevent deterioration and degradation of the seal over time.
Suggestions for Future Research

To further improve the knowledge regarding the effect of the direction
of finishing and polishing procedures on marginal adaptation of resin-based
134
composite, different types of adhesive should be assessed. Indeed, this study
used only one type of adhesive which was a three step total-etch. With the
variety of adhesives present on the market and used in dentistry, the effect of
high and low viscosity adhesives as well as self-etch adhesives should be
evaluated.
Because this study has been conducted in vitro, conclusions cannot
necessarily be extended and generalized to clinical situations. A similar
design of quantitative margin analysis conducted in vivo would be relevant.
135
CHAPTER VI
CONCLUSION
This study compared the effect on marginal adaptation of polishing

from the resin-based composite to the tooth structure (C-T) versus polishing
from the tooth structure to the resin-based composite (T-C), using two types
of materials: a nanofilled composite (FSU, 3M ESPE) and a microhybrid
(PT4, Kerr), and two polishing systems: Sof-Lex XT discs (3M ESPE) and
rubber polishing cups (HiLuster, Kerr). The results of the present study
revealed that:
A significant difference exists between polishing directions. Therefore,

the main null hypothesis (1) was rejected.
A significant difference exists between polishing techniques SL and R.

Therefore, the main null hypothesis (2) was rejected.
A significant difference exists between resin-based composites FSU

and PT4. Therefore, the main null hypothesis (3) was rejected.
Within the limitations of this study, the following conclusions were
drawn:
Polishing from the resin-based composite to the tooth structure (C-T)

leads to better marginal adaptation than polishing from the tooth
structure to the resin-based composite (T-C).
136
The use of Sof-Lex discs produces more marginal irregularities than

rubber polishers when used from resin-based composite to tooth
structure (C-T) with Filtek Supreme Ultra (FSU).
The use of Point4 (PT4) resin-based composite results in more

continuous margins than Filtek Supreme Ultra (FSU) when polished
from resin-based composite to tooth structure (C-T).
The use of Filtek Supreme Ultra (FSU) resin-based composite produces

more marginal gaps when polished from tooth structure to resin-based
composite (T-C).
137
REFERENCES
3M ESPE. Filtek Supreme Ultra Universal Restorative. Available at:

http://www.everyonehasashade.com/main/pdf/IFU_supreme_ultra_v3.
pdf. Accessed July, 2011.
3M ESPE. Filtek Supreme Ultra Universal Restorative System - Technical
Product Profile. Available at:
http://www.everyonehasashade.com/main/pdf/filtek_technical_produc
t_profile_final.pdf2011.
Albers HF (2002) Tooth-colored restoratives: principles and techniques. 9th
ed. Hamilton, Ont. ; Lewiston, NY: BC Decker.
Ardu S, Braut V, Uhac I, Benbachir N, Feilzer AJ, Krejci I. (2009). Influence
of mechanical and chemical degradation on surface gloss of resin
composite materials. Am J Dent 22(5):264-268.
Asmussen E, Peutzfeldt A. (1998). Influence of UEDMA BisGMA and
TEGDMA on selected mechanical properties of experimental resin
composites. Dent Mater 14(1):51-56.
Asmussen E, Jorgensen KD. (1972). A microscopic investigation of the
adaptation of some plastic filling materials to dental cavity walls.
Acta Odontol Scand 30(1):3-21.
Bagheri M, Ghavamnasiri M. (2008). Effect of cavosurface margin
configuration of Class V cavity preparations on microleakage of
composite resin restorations. J Contemp Dent Pract 9(2):122-129.
Baratieri L. N., Monteiro S.J., Caldeira de Andrada M. A., Cardoso L. C,
Ritter A. V. (1998). Esthetics : direct adhesive restorations on
fractured anterior teeth.
Baratieri LN, Ritter AV. (2005). Critical appraisal. To bevel or not in anterior
composites. J Esthet Restor Dent 17(4):264-269.
Barghi N, Lind SD. (2000). A guide to polishing direct composite resin
restorations. Compend Contin Educ Dent 21(2):138-42, 144.
Barkmeier WW, Shaffer SE, Gwinnett AJ. (1986). Effects of 15 vs 60 second
enamel acid conditioning on adhesion and morphology. Oper Dent
11(3):111-116.
138
Bayne SC, Taylor DF, Heymann HO. (1992). Protection hypothesis for
composite wear. Dent Mater 8(5):305-309.
Bergenholtz G, Cox CF, Loesche WJ, Syed SA. (1982). Bacterial leakage
around dental restorations: its effect on the dental pulp. J Oral
Pathol 11(6):439-450.
Berger SB, Palialol AR, Cavalli V, Giannini M. (2011). Surface roughness
and staining susceptibility of composite resins after finishing and
polishing. J Esthet Restor Dent 23(1):34-43.
Berger SB, Palialol AR, Cavalli V, Giannini M. (2009). Characterization of
water sorption, solubility and filler particles of light-cured composite
resins. Braz Dent J 20(4):314-318.
Bernardo M, Luis H, Martin MD, Leroux BG, Rue T, Leitao J, et al. (2007).
Survival and reasons for failure of amalgam versus composite
posterior restorations placed in a randomized clinical trial. J Am
Dent Assoc 138(6):775-783.
Blunck U, Zaslansky P. (2011). Enamel margin integrity of Class I one-bottle
all-in-one adhesives-based restorations. J Adhes Dent 13(1):23-29.
Blunck U, Roulet JF. (1989). In vitro marginal quality of dentin-bonded
composite resins in Class V cavities. Quintessence Int 20(6):407-412.
Bowen RL. (1956). Use of epoxy resins in restorative materials. J Dent Res
35(3):360-369.
Brackett WW, Gilpatrick RO, Gunnin TD. (1997). Effect of finishing method
on the microleakage of Class V resin composite restorations. Am J
Dent 10(4):189-191.
Braem M, Finger W, Van Doren VE, Lambrechts P, Vanherle G. (1989).
Mechanical properties and filler fraction of dental composites. Dent
Mater 5(5):346-348.
Brannstrom M. (1984). Communication between the oral cavity and the
dental pulp associated with restorative treatment. Oper Dent 9(2):5768.
Buonocore MG. (1955). A simple method of increasing the adhesion of acrylic
filling materials to enamel surfaces. J Dent Res 34(6):849-853.
Buonocore MG, Matsui A, Gwinnett AJ. (1968). Penetration of resin dental
materials into enamel surfaces with reference to bonding. Arch Oral
Biol 13(1):61-70.
139
Cenci MS, Pereira-Cenci T, Cury JA, ten Cate JM. (2009). Relationship
between gap size and dentine secondary caries formation assessed in
a microcosm biofilm model. Caries Res 43(2):97-102.
Cenci MS, Tenuta LMA, Pereira-Cenci T, Del Bel Cury AA, ten Cate JM,
Cury JA. (2008). Effect of microleakage and fluoride on enameldentine demineralization around restorations. Caries Res 42(5):369379.
Cenci MS, Venturini D, Pereira-Cenci T, Piva E, Demarco FF. (2008). The
effect of polishing techniques and time on the surface characteristics
and sealing ability of resin composite restorations after one-year
storage. Oper Dent 33(2):169-176.
Crim GA, Swartz ML, Phillips RW. (1984). An evaluation of cavosurface
design and microleakage. Gen Dent 32(1):56-58.
Davidson CL, Duysters PP, De Lange C, Bausch JR. (1981). Structural
changes in composite surface material after dry polishing. J Oral
Rehabil 8(5):431-439.
Davila JM, Gwinnett AJ, Robles JC. (1989). Marginal adaptation of
composite resins and dentinal bonding agents. Future Dent 4(3):1213.
Emilson CG, Krasse B. (1985). Support for and implications of the specific
plaque hypothesis. Scand J Dent Res 93(2):96-104.
Ergucu Z, Turkun LS. (2007). Surface roughness of novel resin composites
polished with one-step systems. Oper Dent 32(2):185-192.
Feilzer AJ, de Gee AJ, Davidson CL. (1990). Relaxation of polymerization
contraction shear stress by hygroscopic expansion. J Dent Res
69(1):36-39.
Feilzer AJ, De Gee AJ, Davidson CL. (1987). Setting stress in composite resin
in relation to configuration of the restoration. J Dent Res
66(11):1636-1639.
Ferracane JL. (2008). Buonocore Lecture. Placing dental composites--a
stressful experience. Oper Dent 33(3):247-257.
Ferracane JL, Marker VA. (1992). Solvent degradation and reduced fracture
toughness in aged composites. J Dent Res 71(1):13-19.
Floyd CJ, Dickens SH. (2006). Network structure of Bis-GMA- and UDMAbased resin systems. Dent Mater 22(12):1143-1149.
140
Forss H, Widstrom E. (2004). Reasons for restorative therapy and the
longevity of restorations in adults. Acta Odontol Scand 62(2):82-86.
Frankenberger R, Lohbauer U, Schaible RB, Nikolaenko SA, Naumann M.
(2008). Luting of ceramic inlays in vitro: marginal quality of self-etch
and etch-and-rinse adhesives versus self-etch cements. Dent Mater
24(2):185-191.
Fukushima M, Setcos JC, Phillips RW. (1988). Marginal fracture of posterior
composite resins. J Am Dent Assoc 117(5):577-583.
Garcia-Godoy F, Krmer N, Feilzer AJ, Frankenberger R. (2010). Long-term
degradation of enamel and dentin bonds: 6-year results in vitro vs. in
vivo. Dent Mater 26(11):1113-1118.
Goncalves F, Kawano Y, Pfeifer C, Stansbury JW, Braga RR. (2009).
Influence of BisGMA, TEGDMA, and BisEMA contents on viscosity,
conversion, and flexural strength of experimental resins and
composites. Eur J Oral Sci 117(4):442-446.
Guggenberger R, Weinmann W. (2000). Exploring beyond methacrylates. Am
J Dent 13(Spec No):82D-84D.
Gwinnett AJ. (1971). Histologic changes in human enamel following
treatment with acidic adhesive conditioning agents. Arch Oral Biol
16(7):731-738.
Jablonski S (1982) Illustrated dictionary of dentistry. Philadelphia:
Saunders.
Jefferies SR. (2007). Abrasive finishing and polishing in restorative
dentistry: a state-of-the-art review. Dent Clin North Am 51(2):379-97,
ix.
Jefferies SR. (1998). The art and science of abrasive finishing and polishing
in restorative dentistry. Dent Clin North Am 42(4):613-627.
Jorgensen KD, Asmussen E. (1978). Occlusal abrasion of a composite
restorative resin with ultra-fine filler--an initial study. Quintessence
Int Dent Dig 9(6):73-78.
Kanemura N, Sano H, Tagami J. (1999). Tensile bond strength to and SEM
evaluation of ground and intact enamel surfaces. J Dent 27(7):523530.
141
Kerr Dental Sybron Dental Specialties. A Guide for using Point 4 Optomized
Particle Composite System. Available at:
http://www.kerrdental.com/index/cms-filesystemaction?file=/kerrdental-products-dfu/dfu-point4.pdf. Accessed July,
2011.
Kidd EA. (1976). Microleakage: a review. J Dent 4(5):199-206.
Kim KH, Ong JL, Okuno O. (2002). The effect of filler loading and
morphology on the mechanical properties of contemporary
composites. J Prosthet Dent 87(6):642-649.
Leinfelder K F, Sockweel C L, Sluder T B. (1982). Two year clinical
evaluation of Profile in posterior teeth. Journal of dental research
61(215, abstract 327).
LeSage BP. (2007). Aesthetic anterior composite restorations: a guide to
direct placement. Dent Clin North Am 51(2):359-78, viii.
Lopes GC, Franke M, Maia HP. (2002). Effect of finishing time and
techniques on marginal sealing ability of two composite restorative
materials. J Prosthet Dent 88(1):32-36.
Lopes GC, Thys DG, Klaus P, Oliveira GM, Widmer N. (2007). Enamel acid
etching: a review. Compend Contin Educ Dent 28(1):18-24; quiz 25,
42.
Maresca C, Pimenta LA, Heymann HO, Ziemiecki TL, Ritter AV. (2010).
Effect of finishing instrumentation on the marginal integrity of resinbased composite restorations. J Esthet Restor Dent 22(2):104-112.
Marghalani HY. (2010). Effect of filler particles on surface roughness of
experimental composite series. J Appl Oral Sci 18(1):59-67.
Marshall GW, Olson LM, Lee CV. (1975). SEM investigation of the
variability of enamel surfaces after simulated clinical acid etching for
pit and fissure sealants. J Dent Res 54(6):1222-1231.
Michalek SM, Shiota T, Ikeda T, Navia JM, McGhee JR. (1975). Virulence of
Streptococcus mutans: biochemical and pathogenic characteristics of
mutant isolates. Proc Soc Exp Biol Med 150(2):498-502.
MjoMjor IA, Qvist V. (1997). Marginal failures of amalgam and composite
restorations. J Dent 25(1):25-30.
142
National Institute of Health, U.S. American Contributions To The New Age
Of Dental Research. Available at:
www.nlm.nih.gov/hmd/pdf/american.pdf.
Porte A, Lutz F, Lund MR, Swartz ML, Cochran MA. (1984). Cavity designs
for composite resins. Oper Dent 9(2):50-56.
Powers JM, Farah JW. (2010). Technique sensitivity in bonding to enamel
and dentin. Compend Contin Educ Dent 31 Spec No 3:1-8; quiz 9.
Powers JM, Craig RG, Sakaguchi RL (2006) Craig's restorative dental
materials. 2nd ed. St. Louis, Mo. ; London: Elsevier Mosby.
Ripa LW, Gwinnett AJ, Buonocore MG. (1966). The "prismless" outer layer of
deciduous and permanent enamel. Arch Oral Biol 11(1):41-48.
Ritter AV. (2005). Direct resin-based composites: current recommendations
for optimal clinical results. Compend Contin Educ Dent 26(7):481-2,
484-90; quiz 492, 527.
Roulet JF, Reich T, Blunck U, Noack M. (1989). Quantitative margin
analysis in the scanning electron microscope. Scanning Microsc
3(1):147-58; discussion 158-9.
Russell RR. (2009). Changing concepts in caries microbiology. Am J Dent
22(5):304-310.
Sabatini C, Blunck U, Denehy G, Munoz C. (2010). Effect of pre-heated
composites and flowable liners on Class II gingival margin gap
formation. Oper Dent 35(6):663-671.
Sabatini C (2007) Effect of pre-heated composites and flowable liners on class
II gingival margin adaptation (dissertation). University of Iowa.
Sarrett DC. (2007) Prediction of clinical outcomes of a restoration based on in
vivo marginal quality evaluation. J Adhes Dent 9 Suppl 1:117-120.
Scherman L. (1990). Adaptation of composites to enamel: relation to the
appearance of marginal staining. Rev Odontostomatol (Paris)
19(1):21-26.
Schmidlin PR, Gohring TN. (2004). Finishing tooth-colored restorations in
vitro: an index of surface alteration and finish-line destruction. Oper
Dent 29(1):80-86.
143
Schneider LF, Cavalcante LM, Silikas N. (2010). Shrinkage Stresses
Generated during Resin-Composite Applications: A Review. J Dent
Biomech 2010:131630. Epub 2009 Sep 30.
Schneider PM, Messer LB, Douglas WH. (1981). The effect of enamel surface
reduction in vitro on the bonding of composite resin to permanent
human enamel. J Dent Res 60(5):895-900.
Simain F, Rompen E, Heinen E. (2010). Dental biofilms. Rev Med Liege
65(10):569-573.
Skjorland KK. (1973). Plaque accumulation on different dental filling
materials. Scand J Dent Res 81(7):538-542.
Speiser AM, Kahn M. (1977). The etched butt-joint margin. ASDC J Dent
Child 44(1):42-45.
Stoodley P, Wefel J, Gieseke A, Debeer D, von Ohle C. (2008). Biofilm plaque
and hydrodynamic effects on mass transfer, fluoride delivery and
caries. J Am Dent Assoc 139(9):1182-1190.
Strassler HE. (1990). Insights and innovations. J Esthet Dent 2(4):122-123.
Sturdevant CM, Barton RE, Strickland WD, Sockwell CL (1985) The Art and
science of operative dentistry. 2nd ed. St. Louis: C.V. Mosby.
Summitt JB (2006) Fundamentals of operative dentistry : a contemporary
approach. 3rd ed. Chicago: Quintessence Pub.
Swift EJ,Jr. (2010). Critical appraisal. Options for dentin/enamel bonding:
Part I. J Esthet Restor Dent 22(1):72-77.
Taylor MJ, Lynch E. (1993). Marginal adaptation. J Dent 21(5):265-273.
Taylor MJ, Lynch E. (1992). Microleakage. J Dent 20(1):3-10.
Ten Cate AR (1985) Oral Histology: Development, Structure, and Function.
Toronto, Canada: Mosby.
Totiam P, Gonzalez-Cabezas C, Fontana MR, Zero DT. (2007). A new in vitro
model to study the relationship of gap size and secondary caries.
Caries Res 41(6):467-473.
Turssi CP, Ferracane JL, Vogel K. (2005). Filler features and their effects on
wear and degree of conversion of particulate dental resin composites.
Biomaterials 26(24):4932-4937.
144
Van Noort R, Davis LG. (1984). The surface finish of composite resin
restorative materials. Br Dent J 157(10):360-364.
Venhoven BA, de Gee AJ, Werner A, Davidson CL. (1996). Influence of filler
parameters on the mechanical coherence of dental restorative resin
composites. Biomaterials 17(7):735-740.
Versluis A, Douglas WH, Cross M, Sakaguchi RL. (1996). Does an
incremental filling technique reduce polymerization shrinkage
stresses? J Dent Res 75(3):871-878.
Watanabe H, Khera SC, Vargas MA, Qian F. (2008). Fracture toughness
comparison of six resin composites. Dent Mater 24(3):418-425.
Whittaker DK. (1982). Structural variations in the surface zone of human
tooth enamel observed by scanning electron microscopy. Arch Oral
Biol 27(5):383-392.
Wu W, Toth EE, Moffa JF, Ellison JA. (1984). Subsurface damage layer of in
vivo worn dental composite restorations. J Dent Res 63(5):675-680.
Yalcin F, Korkmaz Y, Baseren M. (2006). The effect of two different polishing
techniques on microleakage of new composites in Class V
restorations. J Contemp Dent Pract 7(5):18-25.
Yap AU, Shah KC, Chew CL. (2003). Marginal gap formation of composites
in dentine: effect of water storage. J Oral Rehabil 30(3):236-242.
Yap AU, Ang HQ, Chong KC. (1998). Influence of finishing time on marginal
sealing ability of new generation composite bonding systems. J Oral
Rehabil 25(11):871-876.
Yu XY, Wieczkowski G, Davis EL, Joynt RB. (1990). Influence of finishing
technique on microleakage. J Esthet Dent 2(5):142-144.
Zalkind MM, Keisar O, Ever-Hadani P, Grinberg R, Sela MN. (1998).
Accumulation of Streptococcus mutans on light-cured composites and
amalgam: an in vitro study. J Esthet Dent 10(4):187-190.

Effect of Finishing and Polishing Direction On The Marginal Adapt

Caricato da

Informazioni sul documento

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Effect of Finishing and Polishing Direction On The Marginal Adapt

Caricato da

Copyright:

Formati disponibili

University of Iowa

Iowa Research Online

Effect of finishing and polishing direction on the

Copyright 2011 Laurie St-Pierre

Follow this and additional works at: http://ir.uiowa.edu/etd

EFFECT OF FINISHING AND POLISHING DIRECTION ON THE

A thesis submitted in partial fulfillment

Thesis Supervisors: Associate Professor Cathia Bergeron

This is to certify that the Masters thesis of

I would like to express my sincere appreciation to my mentors, Dr.

LIST OF TABLES ............................................................................................. vii

Physical and Mechanical Properties ......................................... 40

Difference between Two Types of Resin-Based Composite ............ 93

Table 1. Marginal Quality Criteria .................................................................. 72

Figure 1. Field emission scanning electron microscope image of

Figure 14. Group assignment and number of specimen according to the

Filtek Supreme Ultra (3M ESPE, St-Paul, MN, USA)

Point 4 (Kerr, Orange, CA, USA)

Polishing technique using a series of Sof-Lex discs

Polishing technique using a fine diamond finishing bur followed

Finishing/polishing direction from resin-based composite resin to

Finishing/polishing direction from tooth structure to resin-based

FeSEM: Field emission scanning electron microscope

Scanning electron microscope

Within the last decades, dental patients expectations toward esthetic

communication with the underlying dentin and trigger tubular fluid

Purpose of the Study

Null Hypothesis (1): there is no difference in the marginal adaptation

Null Hypothesis (2): there is no difference in marginal adaptation

Null Hypothesis (3): there is no difference in marginal adaptation

Enamel and Dentin Composition

Figure 1. Field emission scanning electron microscope image of prepared

Figure 2. Classification of dental adhesives.

Because the outer enamel surface is hypermineralized, has high

Formulation and Chemistry

Organic Polymer Matrix

Inorganic Filler Particles

Classification of Resin-Based Composites

Macrofilled resins have large particles of spherical shape ranging from 20 to

Reasons for Failure

Lactobacilli have been shown to be associated with dental caries (Emilson

Pseudomonas aeruginosa, fusobacteria, staphylococcus species, actinomyces

mutans. The authors compared also the accumulation of Streptococcus

statistically higher bacterial accumulation than amalgam samples but no

Effect of Bacterial Leakage around

Finishing and Polishing Techniques

Effect of Finishing and Polishing

adaptation of resin-based composites (Yu et al., 1990; Brackett et al., 1997;

Delayed vs. Immediate Polishing

disregarding the possible effect of finishing and polishing after hygroscopic

considerable amount of heat which can be detrimental to the adhesive bond

Wet vs Dry Finishing

Effect of the Type of Resin-Based Composite on

Physical and Mechanical Properties

Effect of the Type of Resin-Based Composite

restorations in vitro. To assess the outcome, the authors used scanning

Measurement of Marginal Seal

In vivo, marginal adaptation can be evaluated clinically with

Scanning Electron Microscopy

Quantitative Margin Analysis

investigations. Does the orientation of finishing/polishing technique affect

Figure 6. Sections of a molar before preparation and restoration.

Figure 7. Enamel was flattened using silicon carbide grinding paper on a