← Back
Synthetic, characterization and cytotoxic studies of ruthenium complexes with Schiff bases encompassing biologically relevant moieties
European Polymer Journal 130 (2020) 109664
Contents lists available at ScienceDirect
European Polymer Journal
journal homepage: www.elsevier.com/locate/europolj
Influence of the organic matrix composition on the polymerization behavior
and bulk properties of resin composites containing thiourethanefunctionalized fillers
T
Ana Paula Fugolina, Ana Rosa Costab, Emilie Konoa, Eleanor Quirka, Jack L. Ferracanea,
⁎
Carmem S. Pfeifera,
a
b
Department of Restorative Dentistry, Division of Biomaterials and Biomechanics, Oregon Health & Science University, Portland, OR, United States
Department of Restorative Dentistry, Division of Dental Materials, Piracicaba Dental School, UNICAMP, Piracicaba, SP, Brazil
A R T I C LE I N FO
A B S T R A C T
Keywords:
Thiourethane oligomer
Silica filler functionalization
Polymerization kinetics
Polymer properties
IR spectroscopy
Dynamic mechanical analysis
Objectives: The incorporation of thiourethane-based oligomeric additives into resin composite formulations leads
to improvement in mechanical properties and reduction in polymerization stress, but may increase viscosity. The
objective of this study was to functionalize filler particle surfaces with thiourethane silane molecules and determine the impact of the inorganic filler loading and surface treatment on the behavior of experimental resin
composites with systematically-varied organic matrices.
Methods: Thiourethane oligomer was synthesized de novo, and grafted to the surface of 0.7um barium glass.
BisGMA and TEGDMA (BT) were combined (at 30:70, 50:50 or 70:30 wt%) to 50 or 75 wt% of methacrylate
(MA-Sil – control) or thiourethane-silanized (TU-Sil) particles. Composites were made polymerizable by the
addition of 0.2 wt% BAPO and 0.05 wt% BHT was added as inhibitor. A mercury arc lamp (320–500 nm) at
800 mW/cm2 was used for all curing procedures. Kinetics of polymerization was assessed by near-IR spectroscopy in real time. Polymerization stress was determined with a cantilever system in real time (Bioman). Flexural
modulus and strength were determined in 3-point bending (25 × 2 × 2 mm). Water sorption and solubility and
film thickness were tested according to ISO 4049. Polymeric network characteristics were analyzed by dynamic
mechanical analysis (DMA). Data was analyzed with two-way ANOVA/Tukey’s test (95%).
Results: Viscosity increased with the increase in BisGMA and/or filler amounts. Overall, TU-Sil containing
composites showed delayed vitrification and higher final DC. Filler concentration did not affect DC neither
flexural strength. DC decreased with increasing BisGMA content. Polymerization stress reduced and flexural
modulus increased for higher filler content, especially for formulations containing TU-Sil particles. The water
stability was positively affected by the increase in amount of BisGMA and inorganic filler particles. In terms of
polymeric network, the addition of TU-Sil particles increased the Tg and decreased the E′ and cross-link density.
Conclusions: With the exception of flexural modulus, all tested properties were significantly impacted by the
matrix viscosity and/or the addition of TU-Sil filler particles. In general, the use of thiourethane oligomers as a
silane coupling agent was able to reinforce the materials and reduce the polymerization stress without negatively
affecting the viscosity of the system.
1. Introduction
Manufacturers place great importance on the rheological properties
of commercial resin composites, because the viscosity may strongly
influence many aspects of their handling characteristics, i.e. ease of
placement, ability to shape, adaptation to the dental substrate, stickiness, and slumping resistance [1]. The viscosity of the organic matrix
also directly affects the amount of inorganic filler that can be added. In
⁎
addition, the viscosity plays an even more complex role on a molecular
level, and affects the onset of diffusional limitations to polymerization,
which in turn affect the final conversion and potentially the degree of
crosslinking of the material [2]. Combined, all of these factors influence
mechanical properties, esthetic characteristics, biocompatibility, and
the overall stability of the dental restoration.
The viscosity of the composite paste is determined by the composition of the organic matrix and the relative amount of inorganic fillers.
Corresponding author at: 2730 SW Moody Avenue, 97201 Portland, OR, United States.
E-mail address: pfeiferc@ohsu.edu (C.S. Pfeifer).
https://doi.org/10.1016/j.eurpolymj.2020.109664
Received 24 January 2020; Received in revised form 23 March 2020; Accepted 3 April 2020
Available online 06 April 2020
0014-3057/ © 2020 Elsevier Ltd. All rights reserved.
European Polymer Journal 130 (2020) 109664
A.P. Fugolin, et al.
Fig. 1. (A) Schematic representation of the thiourethane silane synthesized by the reaction of a trifunctional thiol, a di-functional isocyanate, and an isocyanatefunctionalized triethoxy silane. The final product bears pendant thiols (-SH) to undergo chain transfer reaction with the methacrylate organic matrix, and pendant
ethoxy silanes to graft to the inorganic filler particle surface (represented in green). The molecular weight is approximately 5 kDa [20]. (B) Chemical structure of the
commercially-sourced 3-(Trimethoxysilyl)propyl methacrylate used on methacrylate-silanized filler particles. (For interpretation of the references to colour in this
figure legend, the reader is referred to the web version of this article.)
has been proposed [18,19], and one study showed that the addition of
50 wt% of thiourethane-functionalized filler particles (TU-Sil) into an
organic matrix composed by BisGMA, UDMA and TEGDMA (50:30:20,
respectively) led to 15–33% polymerization stress reduction and up to
1.5 fold increase in fracture toughness, without significantly affecting
the viscosity [19]. In order to take advantage of the addition of these
stress-reducing/reinforcing agents, while still producing mixtures containing high volume fractions of inorganic fillers, it is crucial to evaluate the impact the different filler treatments have on the viscosity of
the material when associated with organic matrices of different initial
viscosities. With this in mind, the present study aimed at functionalizing filler particles with oligomeric thiourethanes and determining the
impact of the inorganic filler content and organic matrix viscosity on
polymerization kinetics, mechanical properties, polymerization stress,
film thickness, water stability, and polymer network characteristics by
dynamic mechanical analysis. The tested hypotheses were: (1) the addition of TU-Sil filler particles will decrease the polymerization stress
without compromising kinetics of polymerization, mechanical properties, viscosity and network formation, and (2) more viscous systems will
present lower rates of polymerization, degree of conversion and polymerization stress.
The most common approach to adjust the viscosity of the organic matrix is to combine monomers of different viscosities, which are influenced by their molecular weight and strength of their secondary intermolecular interactions [3]. It has been shown that the combination
of bisphenol A-glycidyl methacrylate (BisGMA) with triethyleneglycol
dimethacrylate (TEGDMA) at mass ratios varying from 50 to 75%
provides increased reactivity, overall conversion and mechanical
properties [3–5]. BisGMA is a high viscosity monomer (> 1000 Pa s)
bearing a rigid aromatic core, with MW = 512 g mol−1 and strong
hydrogen bonding donor potential provided by two hydroxyl groups
per molecule. TEGDMA is a flexible, low viscosity (0.05 Pa s) aliphatic
diluent monomer with MW = 286 g mol−1, and limited hydrogen
bonding acceptor potential [5]. The final viscosity is determined by the
amount of the inorganic filler incorporated into the organic matrix, i.e.
increasing the filler fraction from 40 vol% to 60 vol% may lead to more
than one order of magnitude increase in viscosity (from 100 to
1000 Pa s) [6], and the size and dispersion of filler particles into the
matrix [1].
Clinical evidence shows that dental composite restorations have
shorter service lives than desirable, often being replaced due to secondary caries [7,8]. Even though commercial low-shrink/low-stress
materials have not shown better performance than conventional materials [9], in vitro evidence suggests that the stress generated during
the polymerization reaction, and the consequent gaps formed at the
composite/tooth interface [10], might increase bacterial colonization
and demineralization at the tooth-restoration interface [11]. Therefore,
the incorporation of stress-reducing additives into the resin composite
formulations has been suggested as in an attempt to overcome this
issue. The broad field of additives includes nanotubes, whiskers, POSS
(polyhedral oligomeric silsesquioxanes), nanogels, PPF (pre-polymerized fillers), thiol-ene [12] and thiourethane oligomers [13–15]. An
important consideration related to the incorporation of such additives
into the resin composite formulation is the increased viscosity [16,17].
For example, while the addition of thiourethane oligomers to the
monomer matrix produced composites with enhanced cure and fracture
toughness and reduced contraction stress, the viscosity of the resin
matrix was increased substantially [16]. This led to investigations into
producing materials with the benefits of the thiourethane, but without
affecting flow and handling properties.
Recently, the use of toughening, stress-reducing thiourethane oligomers as a surface modifier grafted onto the surface of inorganic fillers
2. Materials & methods
2.1. Synthesis of thiourethane and functionalization of barium particles
The thiourethane synthesis has been previously described [19].
Briefly, a trifunctional thiol (trimethylol-tris-3-mercaptopropionate,
TMP),
difunctional
isocyanate
(1-isocyanato-4-[(4-isocyanatocyclohexyl) methyl] cyclohexane, DHDI), and a trimethoxy silane – (3-(triethoxysilyl)propyl isocyanate) were combined at 2.5:1:1
mol% in solution. The solvent for the reaction was methylene chloride
and trimethylamine was used as catalyst. The oligomer was purified by
precipitation in hexanes and roto-evaporation. The oligomer characterization was carried out with mid-IR (by the disappearance of the
isocyanate peak at 2270 cm−1) and NMR spectroscopy (resonance
signals at 3.70 ppm) (Fig. 1).
For the thiourethane filler particles functionalization procedure, the
oligomer was mixed in an ethanol:millipore water solution (80:20 vol
%, respectively) acidified by glacial acetic acid (pH ~ 4.5). The
thiourethane was added at 2 wt% of of the ethanol:water solution. Neat
2
European Polymer Journal 130 (2020) 109664
A.P. Fugolin, et al.
2.5. Polymerization stress
barium (0.7 µm ultrafine, Schott Dental Glass, Landshut, Germany) was
incorporated into the solution and kept under magnetic agitation for
24 h, filtered, washed in hexanes, and dried for 4 days at 37 °C. Filler
silanization efficiency was checked using thermogravimetric analysis
(Discovery TGA55, TA Instruments – Waters LLC, New Castle, DE).
Approximately 15 mg of filler was placed in a high temperature platinum pan and subjected to a heat ramp (50–850 °C, 10 °C/min) and the
mass loss was recorded over time.
Polymerization stress (PS) was tested in real-time for 300 s during
the photopolymerization (800 mW/cm2) on disk-shaped samples (5 mm
diameter × 1 mm thick) in a single cantilever device “Bioman” [21]. In
brief, the system consists of a vertical piston placed perpendicular to the
load cell axis and opposite to a fused silica glass plate maintaining a 1mm gap between them, in which the uncured sample is loaded. In order
to avoid sample debonding during the photopolymerization, the piston
surface was treated with a metal primer (Z-Prime Plus, Bisco Inc,
Schaumburg, IL, USA) and the glass surface with silane (Ceramic
Primer, 3M Oral Care, St. Paul, MN, USA) (n = 5).
2.2. Resin composite preparation
The organic matrix of the experimental resin composites consisted
of BisGMA (Bisphenol A diglycidyl dimethacrylate) and TEGDMA
(triethylene glycol dimethacrylate) (BT) combined at 30:70, 50:50 or
70:30 wt%. The inorganic content was based on thiourethane-functionalized filler particles (Tu-Sil – described above) or its commercially
methacrylate-silanized analogous version (MA-Sil – 0.7 µm Ultrafine,
Schott Dental Glass, Landshut, Germany) incorporated in 50 or 75 wt%,
totalizing 12 experimental groups. The formulations were made polymerizable by the addition of 0.2 wt% BAPO (Phenylbis(2,4,6-trimethylbenzoyl)phosphine oxide) and 0.05 wt% of BHT (2,6-di-tertbutyl-4-methylphenol) was incorporated as inhibitor.
All photocuring procedures were performed with a mercury arc
lamp (EXFO Acticure 4000 UV Cure; Mississauga, Ontario, Canada)
filtered to 320–500 nm and delivering 800 mW/cm2 directly to the
surface of the specimen, already accounting for light loss through the
different test set ups. The irradiance of the arc lamp was characterized
daily prior to starting the experiments using a power meter based on a
thermopile system (Molectron, Portland, OR, USA). The wavelength
distribution was characterized with a bench-top UV–Vis apparatus
(USB2000, Ocean Optics, Largo, FL, USA).
2.6. Flexural Strength, elastic modulus and toughness
Flexural Strength (FS), Flexural Modulus (FE) and Toughness (GIC)
were assessed by three-point bending test according to ISO 4049.
Rectangular bars (2.0 × 2.0 × 25.0 mm) were produced in metal
molds placed between two glass slides and polymerized for 120 s
bottom and surface samples, at a distance of 7 cm between the light
guide tip and the glass slide delivering 800 mW/cm2. This distance
created a light spot size sufficient to cover the entire surface of the
specimen, avoiding multiple expositions and uncontrolled overlapping.
Bars were stored in dry conditions for 24 at room temperature (n = 10).
Specimens were tested at 0.5 mm/min cross-head speed with 20 mm
distance between supports. FS (MPa) and FE modulus (GPa) were calculated according to the following equations:
Three disks (rubber molds of 10 mm diameter and 0.8 mm thickness), laminated between two glass slides, were photoactivated and the
polymerization kinetics was recorded in real-time for 300 s with nearinfrared spectroscopy (2 scans per spectrum with 4 cm−1 resolution;
Nicolet 6700 FTIR Spectrometer, Thermo Electron Corporation,
Waltham, MA, USA –) in real-time during the photopolymerization
(300 s, 4.2 cm distance between the light guide and glass slide surface,
delivering 800 mW/cm2 as measured at the specimen surface). The area
corresponding to the methacrylate double bond overtone at 6165 cm−1
was followed to calculate the degree of conversion and the polymerization rate was obtained from the first derivative of the conversion
versus time curve (n = 3).
SL =
M1 − M 3
× 1000
V
(2)
FE =
3LD3
2wdh3
(4)
The viscosity of the organic matrices were assessed by a cone-plate
rheometer (ARES, TA Instruments, New Castle, DE, USA).
Approximately 1 g of each resin was placed between 20-mm diameter
plates and tested at 1 Hz with a gap of 0.3 mm (n = 3).
The viscosity of the composites was assessed by film thickness (FT).
Sample preparation was performed according to ISO Specification 4049
and as described previously in the literature [22]. Briefly, the 0.2 g of
the uncured composites was sandwiched between two Mylar strips and
pressed between two glass slabs statically loaded with 2 kg. After 60 s,
the load was removed and the samples photocured for 60 s at 800 mW/
cm2. The thickness of the resin composite films were measured using a
digital caliper at three different locations for each sample and the
average of the measurements was considered as the sample final
thickness (n = 3).
The same discs used to follow polymerization kinetics were subject
to water sorption and solubility, according to ISO 4049. The initial mass
(M1) was measured before incubation in 5 ml of millipore water, the
mass after 7-day water storage water incubation (M2), and the final
mass (M3) after the sample mass was stabilized under dry storage in a
desiccator connected to the house vacuum. Water sorption (WS) and
solubility (SL) in µg/mm3 were calculated according to equations (1)
and (2), respectively:
(1)
(3)
2.7. Viscosity and film thickness
2.4. Water sorption and solubility
M 2 − M3
× 1000
V
L × D3
× 10−3
4 × w × h3 × d
where L is the maximum load (N), D the span between the supports
(mm), w the specimen width (mm), h the specimen height (mm), and d
the deflection corresponding to L (mm).
Toughness was calculated as the area under stress-strain curve up to
the fracture load.
2.3. Polymerization kinetics
WS =
FS =
2.8. Dynamic mechanical analysis
Glass transition temperature (Tg), storage modulus (E′), and
polymer cross-link density were assessed on sample beams (15 mm
long × 3 mm wide × 1 mm thick) irradiated for 120 s at 800 mW/cm2.
In order to maximize the degree of conversion and prevent any further
polymerization during the test, the bars were submitted to a heat
treatment at 170 °C for 15 h. While that produces networks that are
different than the ones produced in the other test set ups, this test was
included to provide insight on the rank order of materials in terms of
where V is the volume of the disc specimen (in mm3).
3
European Polymer Journal 130 (2020) 109664
A.P. Fugolin, et al.
rates, with a quasi-plateau in rate between 10 and 30% conversion for
all of the BT 30:70 formulations and in BT 50:50 75% TU-Sil resin
composite (Fig. 2). In relation to DC at RPMAX, in general, the highest
values were attained for BT 30:70, with BT 70:30 having the lowest
conversions though not significantly different than BT50:50.
The results for WS (water sorption) and SL (solubility) are presented
in Fig. 3. The WS values ranged between 33.42 ± 1.59 and
15.38 ± 0.92 µg/mm3 and SL between 7.43 ± 0.92 and
2.65 ± 0.92 µg/mm3. In general, the WS and SL values decreased with
increases in BisGMA and filler loading. In addition, although there is no
significant difference, there is a trend for formulations containing TUSil particles to show lower values of WS and SL than MA-Sil-based resin
composites. There was no correlation between final degree of conversion and SL (R2 = 0.137) or WS (R2 = 0.001).
Polymerization stress as a function of time curves are depicted in
Fig. 4A and the average/standard deviation values in Fig. 4B. In general, composites with higher BisGMA concentration (BT 70:30) showed
reduced PS in relation to BT 30:70 and 50:50 (significantly for 50% MASIL and 75% MA-Sil). Additionally, half of the composites with higher
filler concentration had lower values of polymerization stress (exceptions were BT 30:70 MA-Sil, BT 70:30 MA-Sil, and BT 30:70 TU-Sil. The
addition of 75 wt% TU-Sil inorganic particles resulted in significantly
lower PS results in comparison to the MA-Sil analogous formulations for
BT 30:70 and 50:50 formulations (reduction of 28 and 26%, respectively). Regarding the viscosities for the resins only, all groups were
statistically different (p < 0.001), with values of in 0.04 ± 0.01,
0.17 ± 0.01, and 1.17 ± 0.02 Pa s for BT 30:70, 50:50 and 70:30,
respectively.
Film thickness (FT) values ranged between 2.47 ± 0.03 and
0.15 ± 0.07 mm (Fig. 4C). The FT increased significantly with the
filler loading and BisGMA content (p < 0.0001). The highest values
were obtained for BT 70:30 75% TU-Sil (2.47 ± 0.03 mm), followed
by BT 70:30 75% MA-Sil (1.6 ± 0.07 mm), BT 50:50 75% TU-Sil
(0.94 ± 0.09 mm), and 50:50 75% MA-Sil (0.6 ± 0.03 mm), and the
TU-Sil versions produced thicker film than the MA-Sil versions.
The mechanical properties results are shown in Fig. 5. In terms of FS
(flexural strength), there is significant difference only between BT
30:70 75% TU-Sil/ BT 50:50 50% TU-Sil / BT 50:50 75% MA-Sil
(121.0 ± 14.9, 123.2 ± 14.8 and 127.5 ± 11.9 MPa, respectively)
and BT 50:50 50% MA-Sil / BT 70:30 50% MA-Sil (89.8 ± 14.9 and
86.7 ± 11.8 MPa, respectively). All other groups showed intermediate
results. E (flexural modulus) was significantly affected by the filler
particle percentage. All resin composites filled at 75 wt% showed
higher values than their analogous versions filled at 50 wt%
(p < 0.05). In terms of GIC, in general, systems containing 50% TU-Sil
showed the highest values followed by 50% MA-Sil compositions
network structure. The bars were tested using a DMA Instrument (DMA
Q800, TA Instruments, New Castle, United States), in tension mode
with the temperature varying from −50 °C to 250 °C, and ramping rate
of 3 °C/min. The Tg was recorded as the peak value of the tan delta
curve and storage modulus (E′) as a function of temperature. For the
calculation of the cross-link density (v), the mean of the E′ in the rubbery plateau was entered into Eq. (5):
v=
E'
3dRT
(5)
where v is the cross-link density (mol/kg), E′ is the storage modulus
(MPa), d is the density (g/ml), R is the gas constant (8.314472 J/mol
K), and T is the temperature (K).
2.9. Statistical analysis
Data were analyzed for normality (Anderson-Darling) and homocedasticity (Bartlett/Levene), then subjected to ANOVA. Three way
ANOVA (organic matrix composition with three levels, filler loading
(%) with two levels, and filler surface treatment with two levels) was
used to determine which factors and interactions were significant.
Subsequently, averages were contrasted by Tukey’s test (α = 0.05).
3. Results
The results of the three-way ANOVA for all tested variables and
their interactions are shown in Table 1. Except for the mechanical
properties in flexure, the triple interaction was not significant for any of
the variables.
Kinetics of polymerization results are presented in Table 2. In terms
of final DC, in general, resin composites containing TU-Sil filler particles reached higher values (ranging from 69.6 to 81.7%) than MA-Sil
systems (ranging from 65.3 to 73.6%), regardless of the filler concentration (with just a few exceptions). The filler loading did not influence conversion within the same filler surface treatment and organic
matrix composition, except in two cases (TU-SIL at 30:70 and 70:30)
where DC was higher with higher filler load,. In general, BT 70:30
showed lower values of conversion (in all cases except for TU-SIL),
whereas BT 30:70 and 50:50 had similar final conversion, regardless of
the filler loading and surface treatment. BT 30:70 ormulations had the
lowest RPMAX values (p < 0.0001), though there were no differences
for the three resins with MA-Sil at 50% filler load. When comparing the
different filler loadings, the highest values were attained by compositions containing 75% filler (BT 50:50 MA-Sil/ TU-Sil and BT 70:30 TUSil – 18.7, 21.6, and 18.4% s−1, respectively). The polymerization rate
as a function of the degree of conversion curves show slow deceleration
Table 1
Three-way ANOVA for all experiments in the present study. The factors were organic matrix (3 levels: BT 30:70, 50:50 and 70:30), percentage of filler particles (2
levels: 50 and 75 wt%), and filler particle surface functionalization (2 levels: MA-Sil and TU-Sil). The signifance level was α = 0.05. Non-significant interactions are
highlighted in bold. Tests.
3-way ANOVA
RPMAX
DC@
RPMAX
Final DC
FS
FE
GIC
WS
SL
PS
FT
Tg
E′
v
Organic Matrix
Filler %
Filler
Treatment
Organic Matrix *
Filler %
Organic Matrix * Filler
Treatment
Filler % * Filler
Treatment
Organic Matrix * Filler % *
Filler Treatment
< 0.0001
< 0.0001
< 0.0001
0.6965
0.0077
0.0131
< 0.0001
0.9780
0.0084
0.0414
< 0.0001
0.3907
0.6270
0.9595
< 0.0001
0.0015
< 0.0001
< 0.0001
< 0.0001
0.1291
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
0.0002
0.0003
< 0.0001
< 0.0001
< 0.0001
0.0022
< 0.0001
< 0.0001
< 0.0001
0.0136
0.0051
< 0.0001
0.0049
0.9859
0.0022
0.0192
0.0426
< 0.0001
< 0.0001
< 0.0001
0.0540
0.0353
0.0108
0.4691
0.0067
0.0933
0.4764
0.2046
0.0003
< 0.0001
< 0.0001
< 0.0001
< 0.0001
0.1535
0.1514
0.0052
0.2227
0.0090
0.1637
0.0337
< 0.0001
< 0.0001
< 0.0001
< 0.0001
< 0.0001
0.0282
0.0010
0.0597
0.5907
0.2121
0.0220
< 0.0001
< 0.0001
0.0001
0.0001
0.1283
0.0017
< 0.0001
0.8284
0.9549
0.1907
0.4332
< 0.0001
< 0.0001
< 0.0001
< 0.0001
4
European Polymer Journal 130 (2020) 109664
A.P. Fugolin, et al.
Table 2
Average and standard deviation of maximum rate of polymerization, degree of conversion at maximum rate of polymerization and final conversion for all tested
formulations. Values followed by different letters within the same variable indicate statistically significant differences (α = 0.05).
BisGMA:TEGDMA (wt%)
Methacrylate or thiourethane functionalized filler content (wt%)
50%
MA-Sil
75%
TU-Sil
MA-Sil
TU-Sil
6.1 (0.6) c
11.2 (0.6) b
11.7 (1.9) b
5.5 (0.3) c
18.7 (0.3) a
11.8 (2.3) b
9.1 (1.9) bc
21.6 (0.3) a
18.4 (2.3) a
26.2 (3.9) ab
16.6 (4.8) bc
11.5 (2.1) c
19.5 (2.1) abc
14.9 (2.2) c
11.8 (2.9) c
27.9 (5.6) a
17.7 (4.5) abc
13.2 (4.2) c
75.4 (0.6) bc
78.7 (1.8) ab
69.6 (1.2) ef
71.6 (1.6) de
70.7 (0.9) de
65.3 (0.8) g
81.7 (0.3) a
81.5 (1.4) a
78.0 (2.5) ab
. −1
Maximum rate of polymerization – RPmax (% s )
30:70
9.5 (2.1) bc
50:50
13.2 (2.7) b
70:30
9.6 (1.9) bc
Degree of conversion at maximum rate of polymerization – Dc at – RPmax (%)
30:70
19.6 (1.2) abc
50:50
15.2 (3.0) c
70:30
13.1 (3.3) c
Final degree of conversion (%)
30:70
73.5 (1.3) cd
50:50
73.6 (0.7) cd
70:30
67.0 (0.2) fg
Fig. 2. Rate of polymerization (%.s−1) as a function of degree of conversion (%) for BisGMA:TEGDMA-based resin composites filled with 0.7 µm MA-Sil or TU-Sil
particles. The polymerization reaction was followed in real time for 300 s as the dental composites were photopolymerized at 800 mW/cm2.
Fig. 3. Water Sorption (WS) and Solubility (SL) results after 7-day water incubation for all tested experimental resin composites. Different letters indicate statistically
significant differences among the groups (α = 0.05).
4. Discussion
(averaged 2.11 ± 0.79 and 1.40 ± 0.41 MPa, respectively). In addition, GIC increased as filler percentage and BisGMA ratio decreased.
Overall, organic matrix and particle surface functionalization did not
affect FS, E or GIC.
Dynamic mechanical analysis results are reported in Table 3. The
samples containing 75 wt% MA-Sil in BT 30:70 and 50:50 matrices did
not survive the test, which made it impossible to analyze the data. Glass
transition temperature (Tg) results ranged from 176.5 ± 3.2 °C to
158.6 ± 1.4 °C, with higher values for systems containing TU-silanized
fillers, especially in 50 wt% formulations. E′ at rubbery plateau was
used to calculate the degree of crossnlinking (υ), which decreased as the
BisGMA ratios increase for the 50%-filled composites. BT 70:30 75%
MA-Sil presented the highest crosslinking density (0.13 ± 0.02 mol/
kg). The TGA results showed weight loss of 12.6% for TU-Sil and 4.2%
for the commercial MA-Sil.
The last 10 years have been marked by the extensive development
of the resin composite organic matrix, including monomers and additives aimed at controlling the polymerization stress and to further the
reinforcing effect provided by the inorganic filler [15,17,20], with the
ultimate goal to increase the lifespan of the dental restorations. In the
case of the incorporation of additives, a recurrent concern is that these
compounds may result in a significant increase in viscosity due to interparticle interactions [17,20]. It is widely known that the initial viscosity
of a polymerizing medium affects kinetics of polymerization, and may
lead to premature gelation, leading to lower final degrees of conversion
[2]. In addition, higher initial viscosity in the organic matrix also
precludes the incorporation of larger amounts of inorganic filler particles. One solution for this issue is the incorporation of additives as a
5
European Polymer Journal 130 (2020) 109664
A.P. Fugolin, et al.
Fig. 4. (A) Polymerization stress (MPa) plotted as a functional of time (seconds) for all formulations grouped by the ratio between BisGMA:TEGDMA. (B) Final
polymerization stress (MPa) after 5 min of photoactivation at 800 mW/cm2. (C) Film thickness (mm) measured in samples of 0.2 g kept under 2 kg static load for 60 s
for all experimental groups. Different letters indicate statistically significant differences among the groups (α = 0.05).
surface treatment of the filler particles, though this, until now, had not
been investigated using a systematic approach in terms of the organic
matrix composition [19,23].
In general, the presence of TU-functionalized fillers did not affect
the maximum rate of polymerization (with one exception, explored in
detail later), nor the degree of conversion at RPMAX. The final degree of
conversion was higher for four out of six formulations containing TU
compared with the MA-functionalized counterparts. The effect of the
different organic matrix compositions was much more marked. Systems
with higher BisGMA concentration (BT 50:50 and BT 70:30) showed
faster polymerization rates (higher RPMAX) than BT 30:70, the less
viscous system. Right at the beginning of the reaction, the high mobility
of the system allows for the propagation constant (kp) to increase,
which leads to the formation of high molecular weight species. As a
result, the system becomes less reactive due to diffusional limitations,
especially of these high molecular weight species. Eventually, the diffusional limitations start to inhibit termination (kt), which leads to
autoacceleration (increase in kp) up to the rate maximum [2]. In formulations with low initial viscosity, autoacceleration happens at much
lower values of conversion. Therefore, the values of RPMAX never increase significantly, and at the same time the termination events also
decrease, since it is increasingly unlikely for termination by combination and/or disproportionation to take place in this reaction medium
[24]. This explains the quasi-plateau observed for the rate of polymerization as a function of degree of conversion for BT 30:70 (Fig. 1),
which indicates that the system was able to maintain enough mobility
to continue propagation at a low rate, leading to higher degree of
conversion values at RPMax. After that point, vitrification started to
develop, with the increase in viscosity limiting propagation, which also
becomes diffusion-controlled. The reaction then starts to decelerate. In
addition to the decrease in viscosity, TEGDMA’s flexible backbone and
lack of strong intermolecular interactions, in contrast to BisGMA, makes
its network less sterically hindered, which also stimulates macro-radical
segmental movement for a longer period of time [25]. Finally, TEGDMA
also has a tendency for cyclization, which contributes to overall conversion, but not to network development and stiffening of the material
[26]. In summary, the addition of higher amounts of TEGDMA into the
formulations shifts the diffusion control of propagation, delaying it’s
onset to higher conversions [25], as denoted by the DC@RPMAX values
found here. It is important to highlight, however, that in spite of the
tendency of TEGDMA to cyclize, this did not affect the crosslinking
density nor the Tg of the materials (Table 3). High levels of cyclization
have been shown to markedly reduce cross-linking density and glass
transition temperature [26–28], but our results demonstrate that at the
concentrations used here, TEGDMA did not jeopardize network formation. This is evidenced by the DMA results, as already mentioned, as
Fig. 5. Flexural Strength (MPa), Elastic Modulus (GPa) and Toughness (MPa) for all resin composites tested in three-point bending after 48 h dry storage. Different
letters indicate statistically significant differences (α = 0.05).
6
European Polymer Journal 130 (2020) 109664
A.P. Fugolin, et al.
Table 3
Average and standard deviation for glass transition temperature (Tg – °C), storage modulus at the rubbery plateau (E′, GPa), and crosslinking density (υ – mol/kg)
results for all formulations obtained from dynamic mechanical analysis test. Values followed by the same letter within the same variable indicate statistical similarity
(α = 0.05).
BisGMA:TEGDMA (wt%)
Methacrylate or thiourethane functionalized filler content (wt%)
50%
Tg (°C)
30:70
50:50
70:30
E′ at the rubbery plateau (GPa)
30:70
50:50
70:30
Cross-link density (υ) (mol/kg)
30:70
50:50
70:30
75%
MA-Sil
TU-Sil
MA-Sil
TU-Sil
165.3 (0.8) bc
163.4 (2.0) c
159.3 (1.3) c
174.4 (1.9) ab
174.3 (10.7) abc
176.5 (3.2) a
NA
NA
169.0 (0.6) abc
158.6 (1.4) c
169.2 (1.2) abc
167.5 (2.0) abc
711.6 (1 1 1) b
602.0 (1 0 0) bc
356.9 (60.0) c
628.5 (59.0) bc
432.6 (68.9) c
349.4 (32.4) c
NA
NA
1476.2 (277.6) a
788.1 (77.8) b
749.6 (87.1) b
615.9 (40.2) bc
0.0597 (0.0093) bc
0.0553 (0.0094) bc
0.0331 (0.0055) c
0.0563 (0.0055) bc
0.0386 (0.0052) c
0.0311 (0.0028) c
NA
NA
0.1339 (0.025) a
0.0732 (0.0070) b
0.0679 (0.0080) b
0.0561 (0.0039) bc
well as by the flexural strength/flexural modulus results. It is important
to highlight that the presence of the thiourethane may have contributed
to prevent TEGDMA cyclization, via possible hydrogen bonding of the
–SH group in TU with the hydrogen-bond acceptor glycol units in
TEGDMA. Finally, it is important to note that the DMA results are obtained in materials with higher conversion overall, achieved after postcure heat treatment. While it is absolutely acknowledged that this does
not directly correlate with the data obtained with the other tests, it does
provide ranking and comparison among the materials tested.
The formulations based on BT 70:30 showed, in general, the lowest
values of DC at RPMAX. The high concentration of BisGMA was indeed
expected to make the reaction environment more hindered, which in
turn led to early autoacceleration of the polymerization reaction [25].
The high initial and increasing viscosity of the reaction environment
also led to limited final DC for this composition [25]. A plateau was also
identified in the polymerization kinetics profile of the BT 50:50 75%
TU-Sil group. This can be explained by the chain-transfer ability of the
pendant thiols with vinyl groups, leading to delayed gelation/vitrification and allowing for increased degree of conversion before the
reaction becomes diffusion-limited. This also explains the increased
final DC presented by all groups containing 75 wt% TU-Sil filler particles [20]. Interestingly, even in the highly hindered BT 70:30 matrix
system, the addition of 75 wt% TU-Sil increased the final DC by nearly
13% in relation to its analogous MA-Sil version (final DC = 78.0 and
65.3%, respectively). This indicates that the chain-transfer reactions at
this higher overall TU concentration delayed the development of diffusion limitations at a level sufficient to overcome the effect of the
initial viscosity, ultimately leading to higher final double bond conversion. This reinforces the crucial role played by the chain-transfer
events in delaying the point in conversion at which the limitation to the
mobility of the reacting species hampers polymerization, which allows
for not only higher final DC, but also for reduced polymerization stress
generation [29].
The same samples used in the kinetics of polymerization test were
subjected to water incubation in order to check water sorption and
solubility of the experimental resin composites. In general, the increase
in BisGMA and increase in TEGDMA ratios made the composites less
susceptible to absorb water. The LogP (octanol-water partition coefficient) of the organic matrices BT 30:70, 50:50 and 70:30 are 2.521,
3.255 and 3.989, respectively. Since higher values of LogP are associated with lower values of water sorption [30], it was expected that the
decrease in the concentration of the hydrophilic TEGDMA and increase
in the hydrophobic BisGMA would make the organic matrices more
hydrophobic (higher LogP) and, ultimately, less water absorbing. Additionally, in general, 75%-filled composites showed lower numeric
values of WS and SL, which is expected based on the reduced volume of
organic matrix. In terms of filler surface treatment, resin composites
containing TU-Sil particles showed lower numeric values of WS and SL,
which may be associated with the hydrophobicity of the thiourethane
oligomer used to functionalize the filler particles in comparison to the
hydrophilic 3-(Trimethoxysilyl)propyl methacrylate. Likewise, in respect to SL, no marked differences were observed among the tested
formulations, irrespective of organic matrix or filler surface compositions.
The filler concentration impacted the stress development, as expected. On one hand, the greater filler content reduces volumetric
shrinkage, but on the other hand, it increases modulus (stiffness) of the
material [31]. This was indeed observed in this study, and these factors
have shown complex interactions with each other and with the degree
of conversion [32]. As for the role of the organic matrix, in general, the
composites with higher BisGMA content (BT 70:30) showed lower stress
values, which is probably correlated to both the lower shrinkage expected to be obtained with the higher molecular weight BisGMA (not
measured here) and the reduced values of final DC imposed by the
hindered reaction environment due to the restriction mobility of the
large and stiff BisGMA molecular structure [2]. At the75 wt% filler
loading level, TU-Sil-functionalized particles led to 28% and 34% lower
stress compared to the MA-Sil counterparts in BT 30:70 and 50:50 organic matrices, respectively. This effect has been shown in previous
studies in methacrylate-based systems containing thiourethanes either
as oligomer or as surface filler particle functionalizing agent
[15,18,19]. When the thiourethane oligomer is added to the organic
matrix, the reduction in polymerization stress is mainly related to the
chain-transfer reactions, which are responsible for delaying the gelation
and vitrification and possibly contributing to stress relaxation. TU also
leads to the formation of a more homogenous polymer network, possibly limiting the development of internal stress, especially after the
diffusion limitation occurs [29]. In the case when the TU is used as a
filler particle coating, a few possible reasons for stress reduction/relief
include: (1) reduced volumetric shrinkage owing to the high molecular
weight TU; (2) reduction in the energy of propagating cracks/stress due
to the low Tg of the oligomer and flexibility of the thiocarbamate bonds
located at the matrix-particle interface; (3) multi-functional, brush-like
structure of the oligomer, providing sites for stress transfer at the interface; (4) possible stress relaxation via dynamic bond formation; (5)
uncoated areas on the surface of TU-Sil particles acting as defect sites,
and finally; (6) greater filler dispersion [33]. In summary, we hypothesize that the same delayed gelation/vitrification demonstrated for
the additives in the matrix is likely at play here, but the fact that this is
now localized at high-stress the matrix-filler interface potentiates the
7
European Polymer Journal 130 (2020) 109664
A.P. Fugolin, et al.
survived the test) showed an approximately 2.5 times greater cross-link
density in comparison with the analogous TU-Sil group. No clear trends
were identified concerning the effect of monomer composition on
crosslinking density, which is due to the fact that competing events are
occurring simultaneously. It is also likely that the differences that may
have existed prior to the thermal treatment were eliminated by the
thermal treatment conducted prior to the DMA test. For example, the
addition of BisGMA increases the modulus due to its backbone rigidity
but also contributes to lower conversion. TEGDMA contributes to network heterogeneity due to is tendency to ciclization, which reduces
properties in flexure and crosslinking, but also makes chain slippage
more difficult, which can potentially increase Tg [42]. Tg values ranged
from 159 to 177 °C, with numerically greater values for TU-containing
materials.
Finally, a clinically-relevant concern is the viscosity (or final
handling characteristics) of the material. In this study, film thickness
measurements were used as a proxy for viscosity [22]. As expected, the
film thickness increased with higher BisGMA concentration and filler
loading. Moreover, a clear trend was observed between increase in film
thickness and decrease in generated stress, which further reinforces the
crucial role played by the initial viscosity on the polymerization reaction.
overall stress reduction, even at an overall lower TU concentration.
It is important to highlight that the stress reduction provided by the
TU-containing materials was achieved while increasing conversion, and
at no expense to the stiffness. In fact, the flexural modulus, strength and
toughness results for some TU-SIL-containing formulations were the
highest recorded. Thiol-containing networks tend to be more homogeneous, in part due to chain-transfer events, and in the specific case of
thiol-isocyanates, the high flexibility and toughness of the covalent
thiocarbamate bonds can also contribute to improved properties [34].
In tandem with the improved filler particle-organic matrix interaction
due to the multifunctional groups available on the thiourethane oligomer, all of these may have enhanced the mechanical behavior. The
addition of higher amounts of inorganic particles did not significantly
impact the flexural strength but increased the flexural modulus results
and decreased the toughness. In the range of filler loading variation
studied here, other investigations have demonstrated either no change
or even a slight decrease in flexural strength with larger amounts of
nanosilica and micro barium silicate [35]. This demonstrates that there
is a threshold for each particle system in which the flexural strength is
positively impacted by the incorporation of higher amounts of inorganic content, and above which hydrodynamic effects (clustering,
jamming and percolating) may actually jeopardize properties [35,36].
The increase in modulus with the increased filler loading was expected,
as previously shown by many others [31,37,38]. The decrease in
toughness in compositions containing higher filler content and BisGMA
ratio was also expected since both factors increase the brittleness of the
systems, which leads to catastrophic failure with little elastic deformation. On the other hand, TEGDMA acts as a flexible crosslinking
agent [39].
Dynamic mechanical analysis was conducted to gain deeper insight
into the network formation and possibly explain the results for properties in bending and polymerization stress. As already noted, it is acknowledged that either set of properties (obtained by DMA or bend
bars) are assessed in distinct polymer networks due to the thermal
treatment prior to the DMA test – this is a limitation of the method, but
still provides useful ranking among the materials. No correlation analysis was attempted between the two sets of data. For BT 30:70 or 50:50
filled with 75% MA-Sil particles, the samples did not survive the test.
This was due to brittleness in the samples after the thermal treatment
applied before the DMA test. This is not completely unexpected since
the methacrylate silane forms rigid covalent bonds with the organic
matrix. The high content of TEGDMA, presumably with more heterogeneous polymer networks due to microgel formation, is also likely a
contributor to the brittleness of the samples. All TU-Sil samples survived, even the ones containing high concentrations of TEGDMA. This
is explained by the greater toughness of TU-containing materials as has
been reported for formulations in which the oligomer was added directly to the monomer matrix [20]. It is interesting to note that the
increased toughness was also observed when the TU was added as a
filler surface functionality. The only statistical differences when comparing analogous methacrylate- and TU-containing pairs was observed
for BT 70:30 loaded at 50 wt%. In this case, the lower conversion observed for the higher BisGMA concentrations highlighted the toughening effect of the TU. In other words, the addition of TU may be a good
alternative to reinforce materials under sub-optimal polymerization
conditions. The E′ at rubbery plateau was used to calculate the degree
of cross-linking of the network above the Tg and below the melting
temperature. The lack of statistical difference among the groups filled at
50 wt% was not expected, but may be related to the post-processing
procedures (samples were kept at 170 °C for 15 h) performed in order to
achieve the highest final DC and prevent any further conversion during
the heating cycle of the test. In fact, a possible decrease in crosslinking
density was expected with the addition of TU [40], which is explained
by the chain-breaking nature of the chain-transfer reactions of thiol to
vinyl [41]. This is actually demonstrated in the 75 wt% filler groups.
The composition containing MA-Sil (BT 70:30 – the only group that
5. Conclusion
The resin composite viscosity played a key role in the kinetics of
polymerization, water sorption and solubility, polymerization stress
and mechanical properties but not on the polymer network formation
assessed by dynamic mechanical analysis test. Despite the reduction in
crosslinking density observed with the addition of thiourethane surfacemodified filler particles, the mechanical properties were either enhanced or not impacted, and the polymerization stress generation was
markedly decreased.
CRediT authorship contribution statement
Ana Paula Fugolin: Methodology, Writing - original draft, Formal
analysis, Supervision, Visualization. Ana Rosa Costa: Investigation,
Data curation, Formal analysis. Emilie Kono: Investigation, Data
curation. Eleanor Quirk: Investigation, Data curation. Jack L.
Ferracane: Writing - review & editing. Carmem S. Pfeifer:
Conceptualization, Methodology, Resources, Project administration,
Funding acquisition, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors acknowledge funding from NIH-NIDCR (U01DE023756; R01-DE026113; K02-DE025280) and support from
Saturday Academy through the Apprenticeships in Science and
Engineering (ASE) program.
References
[1] K. Al-Ahdal, N. Silikas, D.C. Watts, Rheological properties of resin composites according to variations in composition and temperature, Dent. Mater. 30 (5) (2014)
517–524.
[2] G. Odian, Principles of Polymerization, Wiley, 2004.
[3] M.T. Lemon, M.S. Jones, J.W. Stansbury, Hydrogen bonding interactions in methacrylate monomers and polymers, J. Biomed. Mater. Res. Part A 83A (3) (2007)
734–746.
[4] E. Asmussen, Factors affecting the quantity of remaining double bonds in
8
European Polymer Journal 130 (2020) 109664
A.P. Fugolin, et al.
Nanogel-based filler-matrix interphase for polymerization stress reduction, J. Dent.
Res. 98 (7) (2019) 779–785.
[24] V. Miletic, Dental Composite Materials for Direct Restorations, 2018.
[25] L.G. Lovell, J.W. Stansbury, D.C. Syrpes, C.N. Bowman, Effects of composition and
reactivity on the reaction kinetics of dimethacrylate/dimethacrylate copolymerizations, Macromolecules 32 (12) (1999) 3913–3921.
[26] J.E. Elliott, L.G. Lovell, C.N. Bowman, Primary cyclization in the polymerization of
bis-GMA and TEGDMA: a modeling approach to understanding the cure of dental
resins, Dent. Mater. 17 (3) (2001) 221–229.
[27] H.M.J. Boots, R.B. Pandey, Qualitative percolation study of free-radical crosslinking polymerization, Polym. Bull. 11 (5) (1984) 415–420.
[28] K.S. Anseth, C.N. Bowman, Kinetic Gelation model predictions of crosslinked
polymer network microstructure, Chem. Eng. Sci. 49 (14) (1994) 2207–2217.
[29] S. Ye, N.B. Cramer, I.R. Smith, K.R. Voigt, C.N. Bowman, Reaction kinetics and
reduced shrinkage stress of thiol-yne-methacrylate and thiol-yne-acrylate ternary
systems, Macromolecules 44 (23) (2011) 9084–9090.
[30] S.H. Dickens, G.M. Flaim, C.J.E. Floyd, Effects of adhesive, base and diluent
monomers on water sorption and conversion of experimental resins, Dent. Mater.
26 (7) (2010) 675–681.
[31] K. Masouras, N. Silikas, D.C. Watts, Correlation of filler content and elastic properties of resin-composites, Dent. Mater. 24 (7) (2008) 932–939.
[32] C.S. Pfeifer, J.L. Ferracane, R.L. Sakaguchi, R.R. Braga, Factors affecting photopolymerization stress in dental composites, J. Dent. Res. 87 (11) (2008) 1043–1047.
[33] P. Amdjadi, A. Ghasemi, F. Najafi, H. Nojehdehian, Pivotal role of filler/matrix
interface in dental composites, 2017.
[34] A.F. Senyurt, H. Wei, C.E. Hoyle, S.G. Piland, T.E. Gould, Ternary thiol−ene/acrylate photopolymers: effect of acrylate structure on mechanical properties,
Macromolecules 40 (14) (2007) 4901–4909.
[35] H.A. Rodríguez, W.M. Kriven, H. Casanova, Development of mechanical properties
in dental resin composite: Effect of filler size and filler aggregation state, Mater. Sci.
Eng., C 101 (2019) 274–282.
[36] Y. Song, Q. Zheng, Concepts and conflicts in nanoparticles reinforcement to polymers beyond hydrodynamics, Prog. Mater Sci. 84 (2016) 1–58.
[37] K.-H. Kim, J.L. Ong, O. Okuno, The effect of filler loading and morphology on the
mechanical properties of contemporary composites, J. Prosthetic Dent. 87 (6)
(2002) 642–649.
[38] R. Wang, E. Habib, X.X. Zhu, Evaluation of the filler packing structures in dental
resin composites: from theory to practice, Dent. Mater. 34 (7) (2018) 1014–1023.
[39] I. Sideridou, V. Tserki, G. Papanastasiou, Study of water sorption, solubility and
modulus of elasticity of light-cured dimethacrylate-based dental resins,
Biomaterials 24 (4) (2003) 655–665.
[40] M.G. Borges, L.M. Barcelos, M.S. Menezes, C.J. Soares, A.P.P. Fugolin, O. Navarro,
V. Huynh, S.H. Lewis, C.S. Pfeifer, Effect of the addition of thiourethane oligomers
on the sol–gel composition of BisGMA/TEGDMA polymer networks, Dent. Mater.
(2019).
[41] K.A. Berchtold, B. Hacioǧlu, L. Lovell, J. Nie, C.N. Bowman, Using changes in initiation and chain transfer rates to probe the kinetics of cross-linking photopolymerizations: effects of chain length dependent termination, Macromolecules 34
(15) (2001) 5103–5111.
[42] M. Rubinstein, R.H. Colby, Polymer Physics, Oxford University Press, New York,
2003.
restorative resin polymers, Eur. J. Oral Sci. 90 (6) (1982) 490–496.
[5] F. Gonçalves, Y. Kawano, C. Pfeifer, J.W. Stansbury, R.R. Braga, Influence of
BisGMA, TEGDMA, and BisEMA contents on viscosity, conversion, and flexural
strength of experimental resins and composites, Eur. J. Oral Sci. 117 (4) (2009)
442–446.
[6] K.A. Schulze, A.A. Zaman, K.-J.M. Söderholm, Effect of filler fraction on strength,
viscosity and porosity of experimental compomer materials, J. Dent. 31 (6) (2003)
373–382.
[7] F.F. Demarco, K. Collares, F.H. Coelho-De-Souza, M.B. Correa, M.S. Cenci,
R.R. Moraes, N.J.M. Opdam, Anterior composite restorations: a systematic review
on long-term survival and reasons for failure, Dent. Mater. 31 (10) (2015)
1214–1224.
[8] F.F. Demarco, M.B. Corrêa, M.S. Cenci, R.R. Moraes, N.J.M. Opdam, Longevity of
posterior composite restorations: not only a matter of materials, Dent. Mater. 28 (1)
(2012) 87–101.
[9] F.J. Burke, R.J. Crisp, A. James, L. Mackenzie, A. Pal, P. Sands, O. Thompson,
W.M. Palin, Two year clinical evaluation of a low-shrink resin composite material in
UK general dental practices, Dent. Mater. 27 (7) (2011) 622–630.
[10] J.L. Ferracane, Buonocore memorial lecture: placing dental composites – a stressful
experience, Operat. Dent. 33 (3) (2008) 247–257.
[11] D. Khvostenko, S. Salehi, S.E. Naleway, T.J. Hilton, J.L. Ferracane, J.C. Mitchell,
J.J. Kruzic, Cyclic mechanical loading promotes bacterial penetration along composite restoration marginal gaps, Dent. Mater. 31 (6) (2015) 702–710.
[12] J.A. Carioscia, H. Lu, J.W. Stanbury, C.N. Bowman, Thiol-ene oligomers as dental
restorative materials, Dent. Mater. 21 (12) (2005) 1137–1143.
[13] E. Habib, R. Wang, Y. Wang, M. Zhu, X.X. Zhu, Inorganic fillers for dental resin
composites: present and future, ACS Biomater. Sci. Eng. 2 (1) (2016) 1–11.
[14] R.R. Moraes, J.W. Garcia, N.D. Wilson, S.H. Lewis, M.D. Barros, B. Yang,
C.S. Pfeifer, J.W. Stansbury, Improved dental adhesive formulations based on reactive nanogel additives, J. Dent. Res. 91 (2) (2012) 179–184.
[15] A. Bacchi, R.L. Consani, G.C. Martim, C.S. Pfeifer, Thio-urethane oligomers improve
the properties of light-cured resin cements, Dent. Mater. 31 (5) (2015) 565–574.
[16] A. Bacchi, C.S. Pfeifer, Rheological and mechanical properties and interfacial stress
development of composite cements modified with thio-urethane oligomers, Dent.
Mater. 32 (8) (2016) 978–986.
[17] R.R. Moraes, J.W. Garcia, M.D. Barros, S.H. Lewis, C.S. Pfeifer, J. Liu,
J.W. Stansbury, Control of polymerization shrinkage and stress in nanogel-modified
monomer and composite materials, Dent. Mater. 27 (6) (2011) 509–519.
[18] A.L. Faria-e-Silva, A. dos Santos, E.M. Girotto, C.S. Pfeifer, Impact of thiourethane
filler surface functionalization on composite properties, J. Appl. Polym. Sci. 136
(25) (2019) 47687.
[19] A.P. Fugolin, D. Sundfeld, J.L. Ferracane, C.S. Pfeifer, Toughening of dental composites with thiourethane-modified filler interfaces, Sci. Rep. 9 (1) (2019) 2286.
[20] A. Bacchi, A. Dobson, J.L. Ferracane, R. Consani, C.S. Pfeifer, Thio-urethanes improve properties of dual-cured composite cements, J. Dent. Res. 93 (12) (2014)
1320–1325.
[21] D.C. Watts, J.D. Satterthwaite, Axial shrinkage-stress depends upon both C-factor
and composite mass, Dent. Mater. 24 (1) (2008) 1–8.
[22] J. Da Costa, R. McPharlin, T. Hilton, J. Ferracane, Effect of heat on the flow of
commercial composites, Am. J. Dent. 22 (2) (2009) 92–96.
[23] B.M. Fronza, I.Y. Rad, P.K. Shah, M.D. Barros, M. Giannini, J.W. Stansbury,
9