BIOMATERIALS/

RESTORATIVE
DENTISTRY

ASSESSING THE EFFECT OF COMPOSITE FORMULATION ON POLYMERIZATION STRESS

JOHN R. CONDON, B.S.; JACK L. FERRACANE, PH.D.

A B S T R A C T

Background. In this study, the authors Introduction of MSAA produced a 30 percent reduc-
measured the magnitude of the polymerization stress tion in polymerization stress in an experimental com-
of a variety of dental composite materials and posite material.
explored the effect of a novel monomer, a methacry- Conclusions. Composites that contain
lated derivative of styrene-allyl alcohol, or MSAA, in lower levels of inorganic filler particles are less likely
reducing polymerization stress. to produce high levels of polymerization stress
Methods. Eleven commercially available com- during placement. Modifications to traditional com-
posites and a series of experimental composites were posite chemistry can result in materials that produce
evaluated in a mechanical testing machine to mea- lower polymerization stress levels.
sure the maximum stress generated during place- Clinical Implications. The polymeriza-
ment in a confined setting. tion stress produced by dental composite materials
Results. A significant relationship between during light-curing is a leading reason for bond failures
higher filler volume and increased polymerization in adhesive restorations, resulting in postoperative
stress was found among the commercial materials. sensitivity, marginal staining and recurrent caries.

When dental composite is placed and light- ment of 876 composite restorations.6 Secondary
cured, the polymerization reaction is accompanied caries was cited in 25 percent of the cases, while
by shrinkage. In a nonconfined setting, such as poor margins (15 percent) and postoperative sen-
the restoration of an incisal edge, most of the sitivity (6 percent) were also significant factors.
shrinkage is transmitted to the relatively large The relationship between polymerization stress
free-surface area, and, thus, it does not cause any and marginal debonding has been explored to a
significant problems. However, when composite is limited extent. Ferracane and colleagues7 found a
placed in a confined setting, such as a Class I significant correlation between the magnitude of
preparation, less of the polymerization shrinkage polymerization stress and marginal staining in ex
can be expressed at the free surfaces. Because it vivo restored specimens for three light-cured com-
is constrained by its adhesion to the wall of the posites. The influence of some factors associated
cavity, this unresolved polymerization shrinkage with polymerization stress also has been exam-
leads to internal stress, which can exceed the ined in the mouth. Opdam and colleagues8 com-
strength of the bond with the surrounding tooth pared bulk placement with incremental place-
structure and cause the interface to fail.1 The ment of one composite with a technique in which
resulting marginal gap can lead to postoperative composite was placed in 48 teeth that were desig-
sensitivity and may provide a site for recurrent nated for extraction for orthodontic reasons.
caries to develop. After a few weeks, the Class I restorations were
Secondary caries has been cited as the most tested for sensitivity and the teeth were extracted
common cause of failure for dental composite and subjected to scanning electron microscopic
restorations.2-5 A survey of 22 dental practitioners analysis and staining to test for microleakage.
in Great Britain reported the reasons for replace- These investigators found higher sensitivity on

JADA, Vol. 131, April 2000 497

Copyright ©1998-2001 American Dental Association. All rights reserved.
BIOMATERIALS/RESTORATIVE DENTISTRY

studies.13,14 Also, although many

proposals that involve control-
Glass Stubs (2) To Actuator
ling the direction, intensity and
Composite duration of light-curing have
Specimen been explored, no clear con-
sensus about a truly advanta-
geous approach has been
reached.15-18 Recently,
researchers examining dental
materials have devoted greater
attention to the effect of the
Noncontact composite formulation itself.
Transducer What are the compositional fac-
tors to look for in a composite
Light-Curing Unit (2) that indicate it will not generate
high levels of stress when
Acrylic Sleeve
To Load Cell placed? How can these factors
be manipulated to produce the
Figure 1. Polymerization stress testing configuration for commercial
next generation of improved
composite materials. composites?
A few efforts to control poly-
loading and a higher incidence about composites’ insufficient merization stress by changing
of marginal gap formation for resistance to intraoral wear, composite formulation have
the bulk-placed composite. In polymerization stress has been reported. Reduced levels of
addition, nearly 40 percent of become the leading area of com- photoinitiator were found to
these Class I restorations cause lower shrinkage rates,
demonstrated leakage at the which, by implication, could
As dental composites
enamel margins because of the lead to lower polymerization
high polymerization stresses. have gained wider stress levels.19 Nonbonded
Other studies examining the acceptance for a microfiller particles have been
effect of different light-curing found to produce significant
growing number
and placement regimens have decreases in polymerization
used this technique and demon- of applications, the stress by acting as stress-
strated significant leakage or need to control relieving sites within the com-
marginal gap formation at posite.20 Similarly, self-cured
polymerization
dentin or enamel margins.9,10 composites contain a greater
At least two studies have shrinkage stress has amount of porosities than light-
shown greater marginal leakage become a pressing cured composites and have a
in in vivo restorations compared slower polymerization rate, both
concern.
with in vitro restorations,11,12 of which tend to reduce poly-
suggesting that the effects of merization stress.21
polymerization stress may be posite research in dentistry Gap formation at the inter-
detrimental even when restora- today. As dental composites face between the composite and
tions are placed under con- have gained wider acceptance the surrounding tooth structure
trolled clinical conditions. for a growing number of appli- occurs because the stress caused
Although the direct evaluation cations, the need to control by polymerization exceeds the
of clinical performance is impor- polymerization shrinkage stress strength of the adhesive bond
tant, establishing a relationship has become a pressing concern. between the composite and the
between the composition of the Placement of the composite in tooth.1 Although polymerization
materials used and the mechan- increments gained early shrinkage is relatively easy to
ical properties of interest should widespread acceptance, despite measure, the tendency to pro-
not be overlooked. only limited evidence of a bene- duce polymerization stress is
With concerns diminishing ficial effect in controlled the factor that more fully pre-

498 JADA, Vol. 131, April 2000

Copyright ©1998-2001 American Dental Association. All rights reserved.
 BIOMATERIALS/RESTORATIVE DENTISTRY

dicts the vital clinical aspects of TABLE

composite performance. In this
study, we evaluated an assort- COMMERCIAL COMPOSITES.
ment of commercial materials to
COMPOSITE MANUFAC- TYPE BATCH FILLER VOLUME
determine their polymerization MATERIAL TURER NUMBER (PERCENT)
stress in a confined setting. We
Durafill Heraeus Microfill BE4104 38
also tested a series of experi- VS Kulzer
mental composites containing a
Epic Parkell Microfill 70401 43
novel monomer to evaluate the TMPT
potential for reducing polymer-
Litefil IIA Shofu Microfill 119725 45
ization stress levels.
Helio- Ivoclar- Microfill 823373 48
MATERIALS AND molar Vivadent
METHODS
Fulfil DENTSPLY/ Midifill 070390 53
L.D. Caulk
One of us (J.C.) used a mechan-
ical testing machine to perform Prisma DENTSPLY/ Midifill 920810 55
TPH L.D. Caulk
polymerization stress tests. Sec-
tions of a 5-millimeter–diameter Estelite Tokuyama Midifill EDY500 56

clear glass rod were sandblasted Herculite SDS Kerr Minifill 504539 57
and treated with a ceramic
Charisma Heraeus Minifill 724304 59
primer and light-cured adhesive Kulzer
(Scotchbond MP, 3M Dental
Tetric Ivoclar- Minifill 902514 62
Products). Two of these glass Vivadent
stubs were bonded in the
Bisfil 2 Bisco Self-cure 049125 65
testing machine, one to a fixture
attached to the movable actu-
ator and one to a fixture back to the testing machine. To composite, which was monitored
attached to the force measuring compensate for the small for 30 minutes), and the max-
load cell (Figure 1). amount of deflection that occurs imum force was divided by the
A positioning jig was used to within the load cell, as well as area of the stub to provide a
properly align the upper stub other sources of mechanical mean amount of stress. Three
with the axis of the testing compliance, the testing machine specimens for each of the com-
machine while it was cemented was programmed to maintain mercial materials were tested.
in place with light-cured com- the gap size at 2.5 mm. The composites chosen repre-
posite. The lower stub was then The degree of confinement in sented a cross-section of the
cemented into place while being a polymerization stress test has broad spectrum of composite
pressed into position by the been rated in terms of the con- types (Table). Included were four
upper stub. These opposing sur- figuration factor, or C-factor, microfills (Durafill VS, Heraeus
faces were then treated with which is defined as the ratio of Kulzer; Epic TMPT, Parkell;
adhesive, which was light- the bonded area to the Litefil IIA, Shofu; and
cured, and the 2.5-mm space unbonded area of a volume of Heliomolar, Ivoclar-Vivadent),
between them was filled with shrinking material.22 For this three minifills containing submi-
composite. Light-curing guns test, the C-factor was equal to cron fillers (Tetric, Ivoclar-
(Optilux, Demetron Research 1.0. A small amount of compli- Vivadent; Charisma, Heraeus
Corp.) were used to apply a 60- ance was still present because Kulzer; and Herculite, SDS Kerr)
second light activation from two of the composite coupling and three midifills containing
opposing sides. To maintain a between the glass stubs and the larger-sized fillers (Fulfil,
highly confined setting that steel posts; however, we esti- DENTSPLY/L.D. Caulk; Estelite,
simulated an intracoronal mated its effect to be insignifi- Tokuyama; and Prisma TPH,
restoration, the noncontact cant. The force generated by the DENTSPLY/L.D. Caulk). The
transducer (Kaman Instru- composite as it cured was moni- self-cured composite Bisfil 2 was
ments) was placed parallel to tored for 10 minutes (except for also tested to examine the effect
the specimen to provide feed- Bisfil 2, Bisco, the self-cured of its different curing mode.

JADA, Vol. 131, April 2000 499

Copyright ©1998-2001 American Dental Association. All rights reserved.
BIOMATERIALS/RESTORATIVE DENTISTRY

with an average of six pendant

9
methacrylate groups flanked by
MAXIMUM STRESS (MEGAPASCALS)

aromatic rings. Its addition has

8 been found to yield improved
compressive strength and
7 degree of conversion of
methacrylate groups. The high
6 mobility of the functional ends
could provide a flexible link by
5 which internal stresses could be
resolved within the growing
4 polymer.
In this experiment, six com-
3 posites were formulated by
replacing MSAA with Bis-GMA
2 at the level of 0, 20, 40, 60, 80

Tetric
H erculite
Litefil
II A

Estelite
H eliomolar

Prisma

Charisma
Epic
Durafill

Fulfil
VS

TMPT

TPH

Bisfil 2
and 100 percent in a light-cured
resin. The Bis-GMA/MSAA com-
bination amounted to 50 per-
cent by weight of the resin pre-
COMPOSITE MATERIAL
sent. These thick monomers
Figure 2. Polymerization stress levels of several commercial composite
were diluted with triethylene
materials. Bars connected by a horizontal line are not significantly dif- glycol dimethacrylate, or
ferent according to analysis of variance/Tukey’s test (P < .05). See the TEGDMA, which made up the
table for names of manufacturers.
remaining 50 percent by weight
of the resin phase. Silane-
MAXIMUM STRESS (MEGAPASCALS)

treated filler particles (78 per-

12
cent by weight [62 percent by
volume]) were added to form a
11 composite.
Five specimens of each of
10
these composites were tested.
For these materials, we devel-
9
oped an improved stress-testing
8 method, which allows testing of
more highly confined specimens.
7 The fixture attached to the
actuator contained a slot into
6
which the light-curing gun was
5 situated so that it illuminated
0 20 40 60 80 100 the composite through the glass
stub. The stub was given a pol-
MSAA (PERCENT)
ished end so that it could
Figure 3. Polymerization stress levels of experimental composite mate-
transmit an adequate amount of
rials containing a methacrylated derivative of styrene-allyl alcohol, or irradiance, which was measured
MSAA. Bars connected by a horizontal line are not significantly dif- at 300 milliwatts per centimeter
ferent according to analysis of variance/Tukey’s test (P < .05).
squared. The space between the
We also tested a series of or MSAA, has been proposed as stubs was set to 0.83 mm (C-
experimental composites, which an adjunct to the common factor = 3.0). The force gener-
incorporated a novel, highly bisphenol A glycidyl ated by the composite as it
functional monomer. The dimethacrylate, or Bis-GMA, cured was monitored for 10 min-
methacrylated derivative of monomer.23 The molecule con- utes, and the maximum force
styrene-allyl alcohol copolymer, sists of a carbon chain backbone was divided by the area of the

500 JADA, Vol. 131, April 2000

Copyright ©1998-2001 American Dental Association. All rights reserved.
 BIOMATERIALS/RESTORATIVE DENTISTRY

stub to provide a mean amount

MAXIMUM STRESS (MEGAPASCALS)

of stress. We used analysis of
variance/Tukey’s test to com- 8
pare the results (P < .05). ▲ Light-Cured ● Self-Cured

▲
7
RESULTS ▲

The polymerization stress ▲

6 ▲
values for the commercial com- ▲ ▲ ●
posite materials ranged
between 4 and 7 megapascals 5
▲
(Figure 2). The microfills (Epic ▲
▲
▲
TMPT, Litefil IIA, Durafill VS 4
and Heliomolar) produced sig-
nificantly less stress than the 3
minifills (Herculite and Tetric). 35 40 45 50 55 60 65

In general, we observed no sig-

FILLER VOLUME (PERCENT)
nificant differences in polymer-
ization stress between the
Figure 4. Linear regression analysis of polymerization stress levels on
minifill and midifill composites. inorganic filler content.
As shown in Figure 3, the
polymerization stress values for been sought by polymer several more heavily filled
the experimental material chemists for years. In the mean- materials than those described
demonstrate that a significant time, some minor modifications above (P-50 and Z100, 3M
change occurs with the first in the composites in use today Dental Products; Bisfil P,
increment of MSAA, amounting might yield materials with con- Bisco), but the stress levels they
to a 20 percent reduction in siderable advantages. produced were so high that the
stress. Higher levels of MSAA While few significant differ- specimens broke or debonded
also produced a decrease in ences were found in the poly- during the light-curing phase.
stress over the control material merization stress levels of the The self-cured material (Bisfil
(that is, 100 percent Bis-GMA). commercial materials, we 2) was excluded from this
The stress levels are higher observed a trend toward higher regression analysis in light of
than those recorded for the com- stress among materials with the observation by Feilzer and
mercial materials because of the higher inorganic content. In- colleagues25 that a self-cured
more highly confined test con- deed, a linear regression analy- composite produces less stress
figuration to which these mate- sis between stress and filler than its light-cured counterpart,
rials were subjected. volume (as reported by the man- as we found in this study.
ufacturers) yielded a strong cor- On the basis of our test
DISCUSSION
relation (r2 = .82) (Figure 4). results, it appears that less
The next generation of dental This is somewhat surprising in heavily filled composites might
restorative materials will light of the many differences in be preferable because they tend
address the shortcomings of the filler size, resin type and photo- to produce lower levels of poly-
dental composites and amalgam initiator systems present among merization stress. However,
in use today. Ultimately, the these materials. A strong corre- lower inorganic filler content
wear-resistance lacking in lation between filler volume of has been found to result in lower
today’s materials may be pro- commercial composites and fracture toughness26 and wear-
vided in more affordable their elastic modulus or stiff- resistance,27 which are critical to
ceramic inlays, enhanced com- ness has been demonstrated.24 It the performance of restored
posites or, eventually, in a appears that the greater stiff- occlusal surfaces. The lower
tissue-engineered enamel ness of the more heavily filled stress levels of the less heavily
replacement. The polymeriza- materials plays a major role in filled materials can be advanta-
tion stress of composite might determining the amount of poly- geous when restoring nonoc-
be eliminated by a successful merization stress produced. clusal surfaces. This is espe-
low-shrinkage resin, which has In this study, we also tested cially true for Class V

JADA, Vol. 131, April 2000 501

Copyright ©1998-2001 American Dental Association. All rights reserved.
BIOMATERIALS/RESTORATIVE DENTISTRY

restorations, which have only a complex procedure than mea- capacity, as either a form of
thin layer of enamel available suring polymerization plastic deformation or
for adhesion. In addition, the shrinkage. A computer-con- anisotropic shrinkage, has not
lower elastic modulus resulting trolled testing machine is been determined, it has been
from the lower filler volume may needed to provide a configura- found to relieve approximately
be desirable for Class V restora- tion with minimal compliance. 80 percent of the predicted
tions, because these composites We used testing configurations stress for a chemical-cured com-
will strain more with the teeth that could compensate for the posite in a setting similar to
under load compared with the major sources of compliance in that used in this research.31
stiffer, more heavily filled com- the load cell and the fixtures. When comparing the results
posites.28 At least one clinical We estimated that the of this study with those of
study exhibited data that sup- remaining source of compliance studies in a clinical setting, we
port this hypothesis.29 For in the system—the thin layer of are hampered by the fact that
occlusal surfaces, material selec- composite bonding the glass often only a limited number of
tion must involve improvements stubs to the steel post—had commercial restorative mate-
that do not compromise the only a minimal effect on the rials are included in a given
wear-resistance of the material. clinical study. In addition, any
The use of a novel monomer The development of prediction of long-term
such as MSAA appears to be marginal integrity also would
one method of reducing the novel monomers and depend on the adhesive bond
polymerization stress of the modified photo- strength, water uptake and
composite, while providing initiation systems as thermal expansion properties of
improved compressive strength the material, as well as its
and degree of conversion of well as the addition resistance to aging. However,
methacrylate groups.23 The of nonbonded the magnitude of composite
mobility of the methacrylate microfiller will lead polymerization stress is an
groups seems to be able to essential, if not dominant,
resolve internal stresses, pos- to a variety of new factor in composite restorative
sibly by producing a structure composites whose performance, and it may pro-
that is more prone to changes in advantages are vide an important predictive
molecular conformation or by measure of clinical success.
redirecting the shrinkage based on reduced The development of novel
toward the free surfaces. The polymerization monomers and modified photo-
MSAA appears to lose its effec- stress. initiation systems as well as the
tiveness at higher concentra- addition of nonbonded micro-
tions because it seems to filler will lead to a variety of
operate as a stress-relieving measured stress values. new composites whose advan-
agent. It is possible that once Using a system without com- tages are based, not on the
the stress has been reduced to a pliance compensation, reduction of shrinkage, but on
certain level, further additions Bouschlicher and colleagues30 reduced polymerization stress.
of MSAA provide little benefit. found polymerization force
CONCLUSION
Completely replacing Bis-GMA values for three light-cured
with MSAA might not be benefi- composites that amounted to In this study, we found that
cial, and its specialized nature stresses in the range of 1 to 5 more heavily filled commercial
may cause it to be non–cost- MPa. The magnitude of the composite materials produce
effective. However, adding low stresses measured typically is significantly higher levels of
levels of MSAA could lead to much less than estimates based polymerization stress than
significant reductions in poly- on final stiffness and shrinkage microfills under constrained
merization stress in various values. This is because of a conditions simulating that of
composites, including heavily stress-relieving mechanism intracoronal restorations. In
filled ones. termed the flow capacity of the addition, the novel monomer
Evaluating polymerization composite. Although the MSAA was found to provide sig-
stress is, in general, a more mechanical nature of the flow nificant reductions in polymer-

502 JADA, Vol. 131, April 2000

Copyright ©1998-2001 American Dental Association. All rights reserved.
 BIOMATERIALS/RESTORATIVE DENTISTRY

ization stress in an experi- 8. Opdam NJ, Feilzer AJ, Roeters JJ, Smale of the polymerization. Biomaterials
I. Class I occlusal composite resin restora- 1996;17:2313-8.
mental composite material. ■ tions: in vivo post-operative sensitivity, wall 20. Condon JR, Ferracane JL. Reduction of
adaptation, and microleakage. Am J Dent composite contraction stress through non-
Mr. Condon is a research associate, Depart- 1998;11:229-34. bonded microfiller particles. Dent Mater
ment of Biomaterials and Biomechanics, 9. Van Dijken JWV, Horstedt P, Waern R. 1998;14:256-60.
School of Dentistry, Oregon Health Sciences Directed polymerization shrinkage versus a 21. Alster D, Feilzer AJ, de Gee AJ, Mol A,
University, 611 S.W. Campus Drive, Portland, horizontal incremental filling technique: Davidson CL. The dependence of shrinkage
Ore. 97201. Address reprint requests to Mr. interfacial adaptation in vivo in Class II cavi- stress reduction on porosity concentration in
Condon. ties. Am J Dent 1998;11:165-72. thin resin layers. J Dent Res 1992;71:1619-22.
10. Opdam NJ, Roeters FJ, Feilzer AJ, Ver- 22. Feilzer AJ, de Gee AJ, Davidson CL.
Dr. Ferracane is a professor and chairman, donschot EH. Marginal integrity and postop- Setting stress in composite resin in relation to
Department of Biomaterials and Biome- erative sensitivity in Class 2 resin composite configuration of the restoration. J Dent Res
chanics, School of Dentistry, Oregon Health restorations in vivo. J Dent 1998;26:555-62. 1987;66:1636-9.
Sciences University, Portland. 11. Abdalla AI, Davidson CL. Comparison of 23. Culbertson BM, Tong Y, Wan Q. Copoly-
the marginal integrity of in vivo and in vitro merization of multi-methacrylate derivatives
This study was supported by research grant Class II restorations. J Dent 1992;21:158-62. of styrene-allyl alcohol copolymer with
DE07079 from the National Institute of 12. Ferrari M, Davidson CL. Sealing perfor- styrene and methyl methacrylate monomers.
Dental and Craniofacial Research, Bethesda, mance of Scotchbond multi-purpose and Z100 J Mater Sci Pure Appl Chem
Md. in Class II restorations. Am J Dent 1997;A34(7):1249-60.
1996;9:145-9. 24. Braem M, Van Doren VE, Lambrechts P,
1. Davidson CL, de Gee AJ, Feilzer A. The 13. Tjan AHL, Berge BH, Lidner C. Effect of Vanherle G. Determination of Young’s mod-
competition between the composite-dentin various incremental techniques on the ulus of dental composites: a phenomenological
bond strength and the polymerization contrac- marginal adaptation of Class II composite model. J Mater Sci 1987;22:2037-42.
tion stress. J Dent Res 1984;63:1396-9. resin restorations. J Prosthet Dent 25. Feilzer AJ, de Gee AJ, Davidson CL.
2. Qvist V, Qvist J, Mjör IA. Placement and 1992;67:62-6. Setting stresses in composites for two dif-
longevity of tooth-colored restorations in Den- 14. Versluis A, Douglas WH, Cross M, ferent curing modes. Dent Mater 1993;9:2-5.
mark. ACTA Odontol Scand 1990;48:305-11. Sakaguchi RL. Does an incremental filling 26. Ferracane JL, Berge HX. Fracture
3. MacInnis WA, Ismail A, Brogan H. Place- technique reduce polymerization shrinkage toughness of experimental dental composites
ment and replacement of restorations in a stresses? J Dent Res 1996;75:871-8. aged in ethanol. J Dent Res 1995;74:1418-23.
military population. J Can Dent Assoc 15. Lutz F, Krejci T, Luescher B, Oldenburg 27. Condon JR, Ferracane JL. In vitro wear
1991;57:227-31. T. Improved proximal margin adaptation of of composite with varied cure, filler level and
4. Mjör IA, Jokstad A. Five-year study of Class II composite resin restorations by use of filler treatment. J Dent Res 1997;76:1405-11.
Class II restorations in permanent teeth light-reflecting wedges. Quintessence Int 28. Tyas MJ. The Class V lesion: etiology
using amalgam, glass polyalkenoate (ionomer) 1986;17:659-64. and restoration. Am J Dent 1995;40:167-70.
cement, and resin-based composite materials. 16. Hilton T, Schwartz R, Ferracane J. 29. Heymann HO, Sturdevant JR, Bayne
J Dent 1993;21:338-43. Microleakage of four Class II resin composite SC, Wilder AD, Slader TB. Examining tooth
5. Friedl KH, Hiller KA, Schalz G. Place- insertion techniques at intraoral temperature. flexure effects. JADA 1991;122:41-7.
ment and replacement of composite restora- Quintessence Int 1997;28:135-44. 30. Bouschlicher MR, Vargas MA, Boyer
tions in Germany. Oper Dent 1995;20:34-8. 17. Mehl A, Hickel R, Kunzelmann KH. DB. Effect of composite type, configuration
6. Wilson NHF, Burke FJT, Mjör IA. Rea- Physical properties and gap formation of factor and laser polymerization on polymer-
sons for placement and replacement of light-cured composites with and without ‘soft- ization contraction forces. Am J Dent
restorations of direct restorative materials by start polymerization.’ J Dent 1997;25:321-30. 1997;10:88-96.
a selected group of practitioners in the United 18. Kanca J, Suh B. Pulse activation: 31. Feilzer AJ, de Gee AJ, Davidson CL.
Kingdom. Quintessence Int 1997;28:245-8. reducing resin-based composite contraction Quantitative determination of stress reduc-
7. Ferracane JL, Condon JR, Pham B, stresses at the enamel cavosurface margins. tion by flow in composite restorations. Dent
Mitchem JC. Relating composite contraction Am J Dent 1999;12:107-12. Mater 1990;6:167-71.
stress to leakage in Class V cavities (abstract 19. Venhoven BAM, De Gee AJ, Davidson
3016). J Dent Res 1999;78:S1. CL. Light initiation of dental resins: dynamics

JADA, Vol. 131, April 2000 503

Copyright ©1998-2001 American Dental Association. All rights reserved.