BisGMA/TEGDMA ratio and filler content effects on shrinkage stress

Journal Identification = DENTAL Article Identification = 1789 Date: April 21, 2011 Time: 2:48 pm

d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) 520–526

avai lab le at www.sc iencedi rec t .com

journa l homepage: www. int l .e lsev ierhea l th .com/ journa ls /dema

BisGMA/TEGDMA ratio and filler content effects onshrinkage stress

Flávia Goncalvesa,∗, Caio L.N. Azevedob, Jack L. Ferracanec, Roberto R. Bragaa

a Department of Dental Materials, School of Dentistry, University of São Paulo, São Paulo, Brazilb Department of Statistics, Institute of Mathematics, Statistics and Computer Science, University of Campinas, Campinas, Brazilc Department of Restorative Dentistry, Division of Biomaterials and Biomechanics, School of Dentistry, Oregon Health & ScienceUniversity, Portland, OR, United States

a r t i c l e i n f o

Article history:

Received 14 September 2010

Accepted 18 January 2011

Keywords:

Resin composites

Organic matrix

Inorganic content

Polymerization stress

Mechanical properties

a b s t r a c t

Objective. To investigate the contributions of BisGMA:TEGDMA and filler content on poly-

merization stress, along with the influence of variables associated with stress development,

namely, degree of conversion, reaction rate, shrinkage, elastic modulus and loss tangent for

a series of experimental dental composites.

Methods. Twenty formulations with BisGMA:TEGDMA ratios of 3:7, 4:6, 5:5, 6:4 and 7:3 and

barium glass filler levels of 40, 50, 60 or 70 wt% were studied. Polymerization stress was deter-

mined in a tensilometer, inserting the composite between acrylic rods fixed to clamps of a

universal test machine and dividing the maximum load recorded by the rods cross-sectional

area. Conversion and reaction rate were determined by infra-red spectroscopy. Shrink-

age was measured by mercury dilatometer. Modulus was obtained by three-point bending.

Loss tangent was determined by dynamic nanoindentation. Regression analyses were per-

formed to estimate the effect of organic and inorganic contents on each studied variable,

while a stepwise forward regression identified significant variables for polymerization

stress.

Results. All variables showed dependence on inorganic concentration and monomeric con-

tent. The resin matrix showed a stronger influence on polymerization stress, conversion

and reaction rate, whereas filler fraction showed a stronger influence on shrinkage, modu-

lus and loss tangent. Shrinkage and conversion were significantly related to polymerization

stress.

Significance. Both the inorganic filler concentration and monomeric content affect poly-

merization stress, but the stronger influence of the resin matrix suggests that it may be

possible to reduce stress by modifying resin composition without sacrificing filler content.

The main challenge is to develop formulations with low shrinkage without sacrificing degree

of conversion.

© 2011 Academy

∗ Corresponding author at: University of São Paulo, School of Dentistr05508-000 São Paulo, SP, Brazil. Tel.: +55 11 3091 7840x224; fax: +55 11 3

E-mail addresses: [email protected], [email protected] (F. Gonca0109-5641/$ – see front matter © 2011 Academy of Dental Materials. Pudoi:10.1016/j.dental.2011.01.007

of Dental Materials. Published by Elsevier Ltd. All rights reserved.

y, Department of Dental Materials, Av. Prof. Lineu Prestes, 2227,091 7840x201.lves).blished by Elsevier Ltd. All rights reserved.

dx.doi.org/10.1016/j.dental.2011.01.007

mailto:[email protected]

mailto:[email protected]



2 7

1

Iabpicdf

tattttabAw[ot

avsridaipf

Gsrlrrbpa

2

2

TBprEidPm

d e n t a l m a t e r i a l s

. Introduction

n recent years, polymerization stress has gained considerablettention from researchers and manufacturers as potentiallyeing one of the main causes for interfacial debonding of com-osite restorations [1]. Though a definite association between

nterfacial gap and secondary caries is yet to be establishedlinically, significant correlations between gap size and lesionepth or mineral loss around composite restorations wereound in vitro [2].

Studies employing commercial materials reported correla-ions between polymerization stress and composite shrinkagend elastic modulus [3–6]. However, it is virtually impossibleo draw conclusions regarding the influence of organic con-ent or interaction effects between matrix and filler contentesting proprietary formulations. Studies using experimen-al materials reported a negative correlation between stressnd base-monomer:diluent ratio and positive correlationsetween stress and shrinkage or degree of conversion [7–10].negative correlation between stress and inorganic contentas verified among composites with similar organic phase

11]. However, no study has systematically evaluated the effectf concomitant variations in both organic and inorganic con-ents.

Reaction rate and viscoelastic behavior have been associ-ted with polymerization stress development, with contro-ersial results. While some studies verified a reduction intress with composites cured at slower rates [12,13], otherseported little to no effect [14,15]. Considering all the variablesnvolved, establishing which factors significantly affect stressevelopment, as well as the influence of composite organicnd inorganic content, would provide helpful information fornterpreting results from commercial materials, as well asroviding guidance for the development of new compositeormulations with low contraction stress.

This study aimed at determining the influence of Bis-MA:TEGDMA ratio and inorganic content on polymerizationtress and related variables, namely, degree of conversion,eaction rate, volumetric shrinkage, flexural modulus, andoss tangent. The response variables showing a significantelationship with polymerization stress were identified byegression analysis. The working hypotheses were that: (1)oth matrix and filler influence the tested variables and (2)olymerization stress is significantly related to the other vari-bles.

. Materials and methods

.1. Composite formulations

wenty formulations were prepared containingisGMA (2,2-bis[4-(2-hydroxy-3-methylacryloxypropoxy)-henyl]propane): TEGDMA (2-methyl-2-propenoic acid) molaratios of 3:7, 4:6, 5:5, 6:4 or 7:3 (ESSTECH Technology Inc.,ssington, PA, USA) and 40, 50, 60 or 70 wt% of silanated bar-
um glass particles with irregular shape (average size: 2 �m;ensity: 2.75 g/cm3, Specialty Glass Products, Willow Grove,A, USA). Camphorquinone and 2-(dimethylamino)ethylethacrylate (Sigma–Aldrich Inc., St. Louis, MO, USA), 2 mol%
( 2 0 1 1 ) 520–526 521

each, were used as photoinitiators. Chemicals were usedas provided by the manufacturers. Composites were mixedby mechanical mixer (SpeedMixer DAC 150 FVZ, HauschildEngineering, Alemanha), in order to avoid the incorporation ofair bubbles and to obtain a homogeneous mixture. Materialswere kept refrigerated until 1 h before testing.

2.2. Polymerization stress

Poly(methyl methacrylate) rods (radius, 3 mm; length, 13 or28 mm) had one of their flat surfaces coarsened with #180-gritsandpaper, primed with a thin layer of methyl methacry-late liquid and then coated with unfilled resin (ScotchbondMultipurpose Plus; 3 M ESPE, St. Paul, MN, USA), whichwas light-cured for 40 s (VIP Junior, Bisco Inc., Schaum-burg, IL, USA). The opposing surface of the shorter rod wasmirror-polished to increase light transmission. The rods wereattached to a testing machine (Instron 5565, Canton, MA,USA), with the short rod inserted into a stainless steel hol-low fixture that allowed the light guide to contact its polishedsurface. The composite (1 mm thick) was inserted betweenthe treated surfaces and light-cured through the short rod.The radiant exposure actually reaching the composite was setat 16 J/cm2(500 mW/cm2 × 32 s). Composite height was mon-itored for 10 min using an extensometer (model 2630-101,Instron) attached to both rods, near the gap with the compos-ite, at a resolution of 0.1 �m. The load cell registered the forcenecessary to maintain the initial specimen height in oppo-sition to the composite shrinkage force. Maximum nominalstress was calculated by dividing the contraction force by thearea of the rod (n = 7).

2.3. Degree of conversion and maximumpolymerization rate (RPmax)

Polymerization was monitored for 10 min using near-infra redspectroscopy (Vertex 70, Bruker Optik GmbH, Germany). Thecomposite was inserted in a silicone matrix (radius, 3.5 mm,height, 0.8 mm) and pressed between two glass slides (invisi-ble to near infra-red radiation). Two spectra (two scans each)per sec were collected, with 4 cm−1 resolution. Three spectrawere obtained prior to light-curing and used as an uncuredreference. The change in area under the 6165 cm−1 absorptionpeak corresponding to the first overtone of the C–H stretchof the methacrylate group was used to calculate the degree ofconversion according to the formula (n = 5):

DC =(

1 − cured

incured

)× 100

For each specimen, RPmax was calculated as the first derivativeof the conversion versus time curve.

2.4. Volumetric shrinkage

A mercury dilatomer was used to determine composite volu-metric shrinkage at 10 min after photopolymerization (ADAHF,
Gaithersburg, MD, USA, n = 3). Approximately 50 mg of com-posite was inserted into a glass tube (radius, 2 mm, height,2 mm) bonded to a glass slide. The glass slide was clampedto a glass column, which was carefully filled with mercury. A


522 d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) 520–526

Table 1 – Averages and standard deviations of polymerization stress (MPa), degree of conversion (%), maximumpolymerization rate (% s−1), volumetric shrinkage (%), elastic modulus (GPa) and loss tangent as a function ofBisGMA:TEGDMA ratio (in mol) and filler content (wt%).

BisGMA:TEGDMA (mol%) Filler content (wt%)

40% 50% 60% 70%

Polymerization stress (MPa)3:7 7.0 (0.4) 7.2 (0.4) 6.0 (0.4) 5.8 (0.4)4:6 6.3 (0.6) 6.2 (0.4) 5.9 (0.3) 5.1 (0.2)5:5 5.8 (0.4) 5.1(0.6) 5.0 (0.5) 4.5 (0.2)6:4 4.8 (0.4) 4.6 (0.4) 4.5(0.1) 4.0 (0.3)7:3 4.3 (0.2) 4.0(0.3) 3.8 (0.2) 3.4 (0.2)

Degree of conversion (%)3:7 80.9(1.4) 79.3 (0.7) 82.0 (0.6) 75.4(0.4)4:6 74.8(1.2) 74.9(1.4) 72.0(1.0) 69.7(0.8)5:5 70.2 (0.9) 76.8 (0.5) 70.4(1.0) 69.7(1.0)6:4 65.1(1.0) 68.2 (0.3) 64.5 (0.4) 61.6(0.8)7:3 63.4(0.7) 61.1 (0.2) 59.7 (0.4) 61.3(0.2)

Maximum polymerization rate (% s−1)3:7 2.1 (0.4) 2.3 (0.4) 3.2 (0.2) 2.2 (0.5)4:6 1.6(0.2) 2.1(0.3) 1.7(0.2) 1.5 (0.3)5:5 1.5(0.1) 2.0 (0.4) 1.7(0.2) 1.4(0.1)6:4 1.3(0.1) 1.7(0.2) 1.3(0.1) 1.1 (0.2)7:3 1.5 (0.2) 1.2(0.1) 1.2(0.1) 1.2(0.1)

Volumetric shrinkage (%)3:7 5.8(0.1) 6.0(0.1) 5.1(0.1) 4.1 (0.1)4:6 6.2(0.3) 5.2(0.1) 4.6(0.1) 3.7 (0.2)5:5 5.1 (0.3) 4.8 (0.4) 3.7(0.1) 2.9(0.1)6:4 4.5 (0.2) 3.9(0.2) 3.0(0.1) 2.6 (0.2)7:3 3.8(0.1) 3.4(0.2) 2.8(0.1) 2.4(0.1)

Flexural modulus (GPa)3:7 1.1 (0.2) 1.8(0.3) 2.8 (0.3) 3.9 (0.4)4:6 1.4(0.2) 2.0(0.3) 2.3 (0.4) 3.8 (0.3)5:5 1.4(0.2) 1.9(0.4) 2.5 (0.2) 3.7 (0.3)6:4 1.5 (0.2) 1.8(0.1) 2.3 (0.3) 3.4 (0.4)7:3 1.4(0.3) 1.4(0.2) 2.3 (0.2) 3.2 (0.4)

Loss tangent3:7 0.130(0.009) 0.091 (0.004) 0.069 (0.006) 0.065 (0.006)

0000

4:6 0.098 (0.003)5:5 0.113(0.009)6:4 0.109(0.006)7:3 0.166(0.005)

linear variable differential transducer probe was placed on themercury surface. Photoactivation was conducted through theglass slide (500 mW/cm2 × 32 s). The probe readings were usedto calculate shrinkage based on the mass and specific grav-ity of the specimen (measured on the cured specimen usingArchimedes method).

2.5. Flexural modulus

Bars (12 mm × 2 mm × 1 mm) were made using a silicone mold(n = 10). Pressure was applied to glass slides over mylar stripscovering the surfaces of the composite that was confinedwithin the mold to ensure smooth surfaces and a consistentthickness in all specimens. Ten minutes after photoactivation
(500 mW/cm2 × 32 s), the specimens were subjected to three-point bending, at 0.5 mm/min crosshead speed (QTest, MTSSystem Corporation, NC, USA). Flexural modulus (E, in GPa)was calculated from data obtained from the initial linear por-
.108(0.003) 0.076 (0.008) 0.066 (0.003)

.102(0.016) 0.099 (0.008) 0.056 (0.005)

.115(0.008) 0.089 (0.008) 0.071 (0.006)

.119(0.002) 0.115(0.009) 0.098 (0.007)

tion of the load-displacement curve, according to the formula:

E = C × L3

4 × b × h3 × d× 10−3

where C is the load (in N) corresponding to the displacement ofthe cross-head, (d, in mm), L is the distance between the sup-ports (10 mm), b is the specimen’s width, and h is its height(both in mm). The small-sized bars were used with the intentof irradiating the entire sample with one exposure of light, soradiant exposure would be standardized. As the intent wasto rank the materials according to their E and correlate theresults with organic and inorganic contents, and not to com-
pare our results with others reported in the literature, weopted for standardizing the radiant exposure, which wouldnot have been possible with the use of longer specimens thatwould have required multiple overlapping exposures.


d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) 520–526 523

0.4 0.5 0.6 0.7

3.5

4.0

4.5

5.0

5.5

6.0

6.5

7.0

7.5P

oly

merization s

tress (

MP

a)

Inorganic content (wt%)

3:7

4:6

5:5

6:4

7:3

0.4 0.5 0.6 0.7

60

65

70

75

80

85

De

gre

e o

f co

nve

rsio

n (

%)


3 : 7 4 : 6 5 : 5 6 : 4 7 : 3

0.4 0.5 0.6 0.7

1.0

1.5

2.0

2.5

3.0

3.5

Ma

xim

um

po

lym

eriza

tio

n r

ate

(%

/s)


3 : 7 4 : 6 5 : 5 6 : 4 7 : 3

0.4 0.5 0.6 0.7

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

Vo

lum

etr

ic s

hrin

ka

ge

(%

)


3 : 7 4 : 6 5 : 5 6 : 4 7 : 3

0.4 0.5 0.6 0.7

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

Ela

stic m

odulu

s (

GP

a)


3 : 7

4 : 6

5 : 5

6 : 4

7 : 3

0.4 0.5 0.6 0.7

0.050

0.075

0.100

0.125

0.150

0.175

Lo

ss t

an

ge

nt


3 : 7 4 : 6 5 : 5 6 : 4 7 : 3

BA

DC

EF

Fig. 1 – Mean and confidence intervals (95%) as a function of inorganic content (wt%) for different BisGMA:TEGDMA molarr maxe

2

A2btSNta1o

atios of polymerization stress (A), degree of conversion (B),lastic modulus (E) and loss tangent (F).

.6. Loss tangent

steel mold was used to make the specimens (radius,.5 mm, height 1 mm, n = 3). The composite was sandwichedetween mylar strips and glass slides to ensure consis-ent thickness, smooth surfaces, and to avoid air inhibition.pecimens were subjected to dynamic nanoindentation (Ub1-anomechanical, Hysitron Inc., Minneapolis, MN, USA). Loss

angent was obtained by applying a 500 �N static load usingBerkovich indenter, followed by a 25 �N dynamic loading at

0 Hz, with amplitude of 1.5 nm. Six indentations were maden each specimen. The test started 10 min after photoactiva-

imum polymerization rate (C), volumetric shrinkage (D),

tion (500 mW/cm2 × 32 s) and required 30 min to run. Since wefound no difference in the loss tangent values as a functionof time, we used the mean value of the six indentations torepresent the loss tangent value of the material.

2.7. Statistical analysis

The generalized linear model theory was used to evaluate the
influence of the quantitative co-variables “BisGMA:TEGDMAratio” and “filler content” on the response variables via regres-sion analysis [16]. The regression model is described below,where VE represents the expected value of the considered


524 d e n t a l m a t e r i a l s 2 7 ( 2 0 1 1 ) 520–526

Table 2 – Estimates of regression analyses coefficients (with 95% confidence intervals) having BisGMA:TEGDMA ratio andfiller content as explanatory variables for polymerization stress, degree of conversion, maximum polymerization rate,volumetric shrinkage, flexural modulus and loss tangent.a

Response variable Coefficient estimate 95% confidence interval

Polymerizationstress

Intercept 14.218 [14.125;14.312]BisGMA:TEGDMA 0.987 [0.987;0.987]Filler content 0.993 [0.993;0.993]

Degree of conversionIntercept 105.301 [101.266;109.496]BisGMA:TEGDMA 0.994 [0.993;0.994]Filler content 0.998 [0.998;0.999]

Maximum polymerizationrate


Volumetricshrinkage


Flexural modulus


Loss tangentIntercept 0.161 [0.129;0.202]BisGMA:TEGDMA 1.008 [1.006;1.011]Filler content 0.983 [0.980;0.986]

variaable ooeffic

4. Discussion

Regression analysis was chosen over a fixed effect model asan attempt to quantify the relative contributions of matrix

Table 3 – Estimates of “stepwise forward” regressionanalysis coefficients (with 95% confidence intervals)having polymerization stress as response variable.a

Coefficient estimate 95%confidenceinterval

Intercept 5.057 [4.901;5.217]Volumetric shrinkage 1.122 [1.067;1.180]Degree of conversion 1.096 [1.042;1.152]

a Coefficients higher than 1 indicate positive effect of the explana-tory variable on the response, while coefficients lower than

a Coefficients higher than 1 indicate positive effect of the explanatoryeffect. Coefficients close to 1 indicate a lesser influence of the variintervals indicates statistically significant difference between the c

variable.

VE = intercept + �1 · [filler content] + �2 · [BisGMA : TEGDMA]

Another regression model was adjusted to evaluate thedependency of polymerization stress on the other responsevariables. Explanatory variables were selected following the“stepwise forward” method using the Akaike’s informationcriterion [16]. In this method, new variables are sequentiallyadded to an initial regression model consisting of only theintercept, with the purpose of identifying a sub-set of signifi-cant explanatory variables [17].

3. Results

Means and standard deviations are displayed in Table 1. Theinteraction plots (Fig. 1) show a general trend for parallelism,but a fair amount of crossover provides evidence for the sig-nificant interaction effect between BisGMA:TEGDMA ratio andfiller content for all the tested variables. Overlapping of theerror bars representing 95% confidence intervals strongly sug-gests lack of statistically significant difference between manygroups.

Table 2 shows the estimates for the regression coefficientsand 95% confidence intervals. Coefficients higher that 1 indi-cate a positive effect on the response variable, and vice versa.Also, coefficients close to 1 indicate a lesser influence of the
variable on the response. Both explanatory variables influ-enced the results of the response variables significantly. Thelack of overlapping between confidence intervals indicatesstatistically significant difference between the coefficients
ble on the response, while coefficients lower than 1 indicate negativen the response. Notice that lack of overlapping between confidenceients of the explanatory variables.

of the explanatory variables. Higher BisGMA:TEGDMA ratiosshowed a stronger negative effect on polymerization stress,conversion and RPmax than higher filler content. Filler contentwas more influential on shrinkage (negative), flexural modu-lus (positive) and loss tangent (negative). Among the responsevariables, conversion and shrinkage were identified as hav-ing significant positive association with polymerization stress,with statistically similar coefficients (Table 3).

1 indicate negative effect. Coefficients close to 1 indicate alesser influence of the variable on the response. Notice that theoverlapping between confidence intervals indicates absence ofstatistically significant difference between coefficients estimates.



2 7

aitomihc

smAtoiFlsto

itidfTapeaitGeawmc

oreewcc

fioGbswObst

a

r

d e n t a l m a t e r i a l s

nd filler on polymerization stress and its determinants. Mostmportantly, regression analysis also was used to identifyhose variables with a significant influence on stress devel-pment. Though regression coefficients are small in value, itust be observed that monomer ratio and filler content varied

n additive increments of 10%. Also, most response variablesave small orders of magnitude, augmenting the impact of theoefficients.

The negative effect of BisGMA:TEGDMA ratio on conver-ion and RPmax has previously been reported [9,18,19]. BisGMAonomer is a high viscosity monomer with a stiff backbone.t high concentrations, it decreases the mobility of the reac-

ion medium to the point that the onset of autoaccelerationccurs at low degrees of conversion. As a consequence, RPmax

s decreased and conversion is halted by vitrification [20,21].iller content also showed a negative effect on both variables,ikely due to a mobility restriction mechanism [22,23]. Regres-ion coefficients associated with filler content were very closeo 1, indicating a very subtle negative effect of this co-variablen conversion and RPmax.

Volumetric shrinkage presented a stronger influence ofnorganic content. The increase in filler content reduceshe concentration of carbon double bonds, therefore reduc-ng shrinkage [11,22]. The increase in BisGMA:TEGDMA ratioecreased shrinkage due to the lower conversion resultingrom a more restricted reaction environment [19,24]. Also,EGDMA-rich matrices show increased shrinkage caused byn antiplasticizing effect, allowing a closer packing of theolymer chains [25]. Flexural modulus also showed a strongerffect of filler content compared to matrix composition, ingreement with previous investigations [8,26,27]. The increasen elastic modulus due to increase in filler generally followshe rule of mixtures [27–29] and, though the highest Bis-MA:TEGDMA ratios were shown to significantly decreaselastic modulus [8,26], most specifically for the filler contentsbove 50% in this study the effect of the resin compositionas overshadowed by the effect of the filler. The lower elasticodulus at the highest BisGMA level is likely due to reduced

onversion and cross-linking.Loss tangent was included in this study as a measurement

f polymer mobility, or damping potential. Loss tangent valuesange between 0 (elastic) and 1 (viscous). The negative influ-nce of the inorganic content, also observed previously [30], isxplained by the reduction in mobility of the polymer network,hile the positive effect of the BisGMA:TEGDMA ratio is asso-

iated with the low elastic modulus resultant from reducedrosslinking and overall conversion [31].

The present study agrees with investigations that veri-ed, separately, the influence of matrix and filler contentn polymerization stress [8,11]. The stronger effect of Bis-MA:TEGDMA ratio than filler content on stress is explainedy the fact that lower ratios increased conversion, RPmax,hrinkage, elastic modulus and decreased loss tangent, all ofhich, theoretically, would have a positive effect on stress.n the other hand, filler content affected stress determinantsoth positively (e.g., elastic modulus) and negatively (conver-
ion, RPmax, shrinkage and loss tangent), possibly offsettingheir effects on stress development.
Among the response variables, only conversion and shrink-ge showed statistically significant positive correlations with

( 2 0 1 1 ) 520–526 525

stress. Positive correlations between conversion or shrinkageand stress have been reported in previous studies [7,9]. Also,the lack of influence of RPmax confirmed its minor relevance inthe presence of other variables [14,32]. However, the absenceof flexural modulus in the regression model is somewhat sur-prising and, at this point, the influence of the material usedas bonding substrate in the testing system must be taken intoconsideration. Though positive correlations between elasticmodulus and stress have been reported with the use of lowas well as high compliance systems [6,9,33], in high compli-ance systems such a relationship is observed only if shrinkageand modulus are directly related for the series of materialstested. In the opposite situation, for instance, for a series ofmaterials with the same organic composition but various fillerlevels, the effect of shrinkage prevails over modulus to definestress [11]. In the present study, subsets displaying a directcorrelation between shrinkage and modulus can be identifiedamong materials with different BisGMA:TEGDMA ratios andsimilar filler content. This relationship seems most evidentabove about 50% filler by weight. Also, shrinkage and modu-lus vary in opposite directions for composites with the sameorganic content and different filler levels. The superimpositionof both situations results in a lack of statistical significance formodulus.

In the present study, the choice of a high compliance sys-tem was based on the results of an in vitro investigation thatverified a better agreement between microleakage data andstress results obtained with the use of acrylic rather than glasssubstrates [33]. The significant correlation found betweenshrinkage and stress in the present study agrees with resultsfrom mathematical models [14] and studies of gap formationin vitro [34], that also reported shrinkage as the most influen-tial factor. Others have argued that high compliance systemsare more representative of the clinical situation where toothflexure or deformation occurs [35].

The stronger influence of the resin matrix on polymeriza-tion stress suggests that it may be possible to reduce stressby modifying resin composition without sacrificing filler con-tent. Also, since reducing shrinkage at the expense of overalldegree of conversion is not advisable, balancing low shrinkageand high conversion is the main challenge.

Acknowledgments

This work is based on a thesis submitted to the grad-uate faculty, University of São Paulo, in partial fulfill-ment of the requirements for the PhD degree in DentalMaterials.

The authors thank ESSTECH for kindly donating themonomers, and CAPES (PDEE 3283-08-9) and FAPESP(07/55476-9) for funding this project.

e f e r e n c e s

[1] Ferracane JL. Buonocore lecture – placing dentalcomposites—a stressful experience. Oper Dent2008;33:247–57.

[2] Cenci MS, Pereira-Cenci T, Cury JA, Ten Cate JM. Relationshipbetween gap size and dentine secondary caries formation



l s 2
526 d e n t a l m a t e r i a
assessed in a microcosm biofilm model. Caries Res2009;43:97–102.

[3] Braga RR, Hilton TJ, Ferracane JL. Contraction stress offlowable composite materials and their efficacy asstress-relieving layers. J Am Dent Assoc 2003;134:721–8.

[4] Cadenaro M, Biasotto M, Scuor N, Breschi L, Davidson CL, DiLenarda R. Assessment of polymerization contraction stressof three composite resins. Dent Mater 2008;24:681–5.

[5] Condon JR, Ferracane JL. Assessing the effect of compositeformulation on polymerization stress. J Am Dent Assoc2000;131:497–503.

[6] Kleverlaan CJ, Feilzer AJ. Polymerization shrinkage andcontraction stress of dental resin composites. Dent Mater2005;21:1150–7.

[7] Feilzer AJ, Dauvillier BS. Effect of TEGDMA/BisGMA ratio onstress development and viscoelastic properties ofexperimental two-paste composites. J Dent Res2003;82:824–8.

[8] Goncalves F, Pfeifer CC, Stansburry JW, Newman SM, BragaR. Influence of matrix composition on polymerization stressdevelopment of experimental composites. Dent Mater2010;26:697–703.

[9] Goncalves F, Pfeifer CS, Ferracane JL, Braga RR. Contractionstress determinants in dimethacrylate composites. J DentRes 2008;87:367–71.

[10] Stansbury JW, Trujillo-Lemon M, Lu H, Ding X, Lin Y, Ge J.Conversion-dependent shrinkage stress and strain in dentalresins and composites. Dent Mater 2005;21:56–67.

[11] Goncalves F, Kawano Y, Braga RR. Contraction stress relatedto composite inorganic content. Dent Mater 2010;26:704–9.

[12] Lim BS, Ferracane JL, Sakaguchi RL, Condon JR. Reduction ofpolymerization contraction stress for dental composites bytwo-step light-activation. Dent Mater 2002;18:436–44.

[13] Witzel MF, Calheiros FC, Goncalves F, Kawano Y, Braga RR.Influence of photoactivation method on conversion,mechanical properties, degradation in ethanol andcontraction stress of resin-based materials. J Dent2005;33:773–9.

[14] Li J, Li H, Fok SL. A mathematical analysis of shrinkagestress development in dental composite restorations duringresin polymerization. Dent Mater 2008;24:923–31.

[15] Pfeifer CS, Ferracane JL, Sakaguchi RL, Braga RR.Photoinitiator content in restorative composites: influenceon degree of conversion, reaction kinetics, volumetricshrinkage and polymerization stress. Am J Dent2009;22:206–10.

[16] Paula GA. Modelos de regressão: com apoio computacional.Versão preliminar. São Paulo: IME-USP; 2004.

[17] Zar JR. Biostatistical analysis. 4th ed. Prentice Hall; 1999.[18] Dickens S, Stansbury J, Choi K, Floyd C. Photopoymerization

kinetics of methacrylate dental resins. Macromolecules2003;36:6043–53.

[19] Floyd CJ, Dickens SH. Network structure of Bis-GMA- andUDMA-based resin systems. Dent Mater 2006;22:1143–9.

7 ( 2 0 1 1 ) 520–526

[20] Lovell LG, Stansbury JW, Syrpes DC, Bowman CN. Effects ofcomposition and reactivity on the reaction kinetics ofdimethacrylate/dimethacrylate copolymerizations.Macromolecules 1999;32:3913–21.

[21] Sideridou I, Tserki V, Papanastasiou G. Effect of chemicalstructure on degree of conversion in light-cureddimethacrylate-based dental resins. Biomaterials2002;23:1819–29.

[22] Garoushi S, Vallittu PK, Watts DC, Lassila LV. Effect ofnanofiller fractions and temperature on polymerizationshrinkage on glass fiber reinforced filling material. DentMater 2008;24:606–10.

[23] Halvorson RH, Erickson RL, Davidson CL. The effect of fillerand silane content on conversion of resin-based composite.Dent Mater 2003;19:327–33.

[24] Dewaele M, Truffier-Boutry D, Devaux J, Leloup G. Volumecontraction in photocured dental resins: theshrinkage-conversion relationship revisited. Dent Mater2006;22:359–65.

[25] Sideridou I, Tserki V, Papanastasiou G. Study of watersorption, solubility and modulus of elasticity of light-cureddimethacrylate-based dental resins. Biomaterials2003;24:655–65.

[26] Asmussen E, Peutzfeldt A. Influence of UEDMA BisGMA andTEGDMA on selected mechanical properties of experimentalresin composites. Dent Mater 1998;14:51–6.

[27] Jones DW, Rizkalla AS. Characterization of experimentalcomposite biomaterials. J Biomed Mater Res B: ApplBiomater 1996;33:89–100.

[28] Braem M, Finger W, Van Doren VE, Lambrechts P, Vanherle G.Mechanical properties and filler fraction of dentalcomposites. Dent Mater 1989;5:346–8.

[29] Masouras K, Silikas N, Watts DC. Correlation of filler contentand elastic properties of resin-composites. Dent Mater2008;24:932–9.

[30] Helvatjoglu-Antoniades M, Papadogiannis Y, Lakes RS,Dionysopoulos P, Papadogiannis D. Dynamic and staticelastic moduli of packable and flowable composite resinsand their development after initial photo curing. Dent Mater2006;22:450–9.

[31] Mesquita RV, Axmann D, Geis-Gerstorfer J. Dynamicvisco-elastic properties of dental composite resins. DentMater 2006;22:258–67.

[32] Pfeifer CS, Ferracane JL, Sakaguchi RL, Braga RR. Factorsaffecting photopolymerization stress in dental composites. JDent Res 2008;87:1043–7.

[33] Boaro LC, Goncalves F, Braga RR. Influence of the bondingsubstrate in dental composite polymerization stress testing.Acta Biomater 2010;6:547–51.

[34] Peutzfeldt A, Asmussen E. Determinants of in vitro gap
formation of resin composites. J Dent 2004;32:109–15.
[35] Lee SH, Chang J, Ferracane J, Lee IB. Influence of instrumentcompliance and specimen thickness on the polymerizationshrinkage stress measurement of light-cured composites.Dent Mater 2007;23:1093–100.


Documents

BisGMA/TEGDMA ratio and filler content effects on shrinkage stress