d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.intl.elsevierhealth.com/journals/dema

The effect of ultra-fast photopolymerisation of

experimental composites on shrinkage stress,
network formation and pulpal temperature rise

Luc D. Randolph a,b,c,∗ , William M. Palin d , David C. Watts e ,

Mathieu Genet a,c , Jacques Devaux b,c , Gaetane Leloup a,b,c,f ,
Julian G. Leprince a,b,c,f
a Louvain Drug Research Institute, Université catholique de Louvain, Brussels, Belgium
b Institute of Condensed Matter and Nanosciences, Bio- and Soft- Matter, Université catholique de Louvain,
Louvain-la-Neuve, Belgium
c CRIBIO (Center for Research and Engineering on Biomaterials), Brussels, Belgium
d Biomaterials Unit, University of Birmingham, College of Medical and Dental Sciences, School of Dentistry, St Chad’s

Queensway, Birmingham, B4 6NN, UK

e School of Dentistry and Photon Science Institute, University of Manchester, UK
f School of Dentistry and Stomatology, Université catholique de Louvain, Brussels, Belgium

a r t i c l e i n f o a b s t r a c t

Article history: Objectives. to complement our previous work by testing the null hypotheses that with short
Received 8 August 2014 curing times and high DC, TPO-based resin composites would exhibit (1) higher polymer-
Received in revised form ization stresses and consequently display (2) higher temperature rise and (3) higher flexural
3 September 2014 modulus, flexural strength and hardness, compared to a conventional CQ-based experimen-
Accepted 3 September 2014 tal composite.
Methods. Two experimental resin composites using either Lucirin-TPO or cam-
phorquinone/DMAEMA as photoinitiators were prepared. Light curing was carried out
Keywords: using spectral outputs adapted to the absorption properties of each initiator. Different irra-
MAPO diation protocols were selected (0.5, 1, 3, 9 s at 500, 1000 and 2000 mW/cm2 for Lucirin-TPO
Experimental based composites and 20 or 40 s at 1000 mW/cm2 for Lucirin-TPO and camphorquinone-
Dental composite based composites). Degree of conversion (DC) was measured in real time by means of FT-NIR
Irradiation parameters spectroscopy. Pulpal temperature rise (T) was studied in a tooth model. Polymerization
Irradiance stress was monitored using the Bioman instrument. For cured specimens, flexural modulus
Polymerization kinetics and flexural strength were determined using a three point bending platform and Vickers
Polymerization stress hardness was determined with a microhardness indentor on samples prior to and after
Tooth model 24 h incubation in 75/25 ethanol/H2 O. Premolars were restored with both materials and
Temperature rise microleakage at the teeth/composite interfaces following restoration was assessed.
Microleakage Results. Lucirin-TPO-based composites irradiated at radiant exposures of 3 J/cm2 and more
exhibited significantly higher DCs, associated with increased flexural moduli and hardness
compared to CQ-based composites. For an ultra-short irradiation time of 1 s at 1000 mW/cm2 ,
TPO-composites displayed similar polymerization stresses compared to CQ-controls with

∗
Corresponding author at: Université catholique de Louvain Louvain Drug Research Institute Avenue E. Mounier 73, B-1200 Brussels
Belgium.
E-mail address: lucrandolph@gmail.com (L.D. Randolph).
http://dx.doi.org/10.1016/j.dental.2014.09.001
0109-5641/© 2014 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289 1281

yet a 25% increase for flexural modulus and 40% increase for hardness measured after
EtOH/H2 O sorption. Higher stress rates were however observed in all curing protocols
compared to CQ-composites. Microleakage was similar between TPO and CQ-composites
irradiated at 1000 mW/cm2 for 3 and 20 s respectively, while a significant increase was
observed for TPO-composites irradiated for 1 s. T measured through a 0.6 mm thick dentin
layer were all below 5.5 ◦ C; TPO-composites exhibited similar or lower values compared to
controls.
Significance. The use of Lucirin-TPO in resin composites along with appropriate curing
conditions may allow for a major reduction of irradiation time while improving mechanical
properties. The amount of stress observed during polymerization in TPO-based composites
can be similar to those using CQ and the cohesion at the restoration-tooth interface was not
affected by short curing times. Contrary to other studies, we found that the temperatures
increases measured during polymerization were all well below the 5.5 ◦ C threshold for the
pulp.
© 2014 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

In a recent paper by our group, using a model experimen-

1. Introduction tal BisGMA/TegDMA composite formulation, we demonstrated
the possibility of improving the degree of conversion (DC)
According to clinical studies, the major reasons for failure
and significantly reducing monomer elution despite a signifi-
of resin composite restorations are secondary caries, restora-
cant reduction of curing time (1 s for a 2 mm composite layer
tion fracture and need for endodontic treatment [1–4]. From
at 1000 mW/cm2 ) [17]. This was achieved first by replacing
a biomimetic standpoint, the materials used should display
the common ketone (camphorquinone)/amine photoinitia-
the same characteristics as the replaced tissues. For dentin,
tion system (CQ) with a Norrish Type I monoacylphosphine
hardness lies in the range of 50–60 HV [5], elastic modulus is
oxide photoinitiator, namely Lucirin-TPO (TPO), and also
around 20–25 GPa [6]. Current commercial resin composites
by choosing appropriate irradiation parameters (wavelength
exhibit flexural moduli and strength in the range 2.4–14.7 GPa
range, irradiance and irradiation time). Several other stud-
and 45–145 MPa, respectively and Vickers hardness falls in
ies have already demonstrated that the curing light source
the range of 18.6–77.7 HV/20 [7]. The wide variation of results
is critical in realizing the potential of TPO and that switch-
may be explained by the different formulations but also the
ing from CQ to TPO, and other photoinitiators with higher
quality of conversion as higher degrees of conversion (DC) usu-
molar absorptivity is advantageous for materials develop-
ally lead to more rigid [8] and harder structures [9]. Besides
ment [18–22]. Further, the clinicians’ desire for reduced
mechanical improvement, the materials should present a reli-
curing time (3 s, or less) has seen an influx to the dental
able interface with the dental tissues (generating minimal
market of high irradiance (>2000 mW/cm2 ) LED light curing
shrinkage stress on cure) and be safe for the pulp tissue (low
units that may not provide sufficient cure of some com-
monomer elution, low temperature rise in the pulp chamber,
posite material types [23,24]. If such short curing times are
etc.). At the same time, resin composites should meet the
desired, there is a requirement for the development of mate-
patients and practitioners requirements, notably the reduc-
rials chemistry, rather than simply increasing the power of
tion of chairside procedures times, for example by shortening
the curing light source. Overall, both intrinsic and extrin-
curing time.
sic parameters should be taken into account in order to
Despite the clear clinical interest of reducing curing time by
optimize the polymerization efficacy and resulting material
improving polymerization kinetics, higher rates could result
properties [25].
in greater stresses at the dentin-composite interface, which
Despite these clear improvements, it remains unclear
may in turn induce cusp deflection and/or gaps [10]. How-
whether the dramatic increase in reaction speed seen in
ever the development of polymerization shrinkage stresses is
TPO-based composites [19,20], associated with improved con-
a complex phenomenon, which was shown to be affected by
version [17,19,20] and mechanical properties [20], will be
multiple factors including not only polymerization rate, but
detrimental in terms of shrinkage stress at short curing times.
also degree of conversion, volumetric shrinkage and elastic
Since the generation of polymerization stress throughout cure
modulus [11]. For example, polymerization stress was shown
is not an intrinsic material property, its comprehensive eval-
to evolve non-linearly with conversion [12] while volumetric
uation requires the investigation of other central parameters,
changes are the combination of both contraction due to the
i.e. degree of conversion, temperature rise and elastic modu-
polymerization and contraction/expansion related to thermal
lus. Hence, the aim of the present study was to complement
effects [13]. Finally, an additional possible concern of faster
our previous work [17] by testing the null hypotheses that
polymerization with higher conversion and higher irradiances
with short curing times and high DC, TPO-based resin com-
is the heat generated during polymerization, which could
posites would exhibit (1) higher polymerization stresses and
damage the underlying pulp [14,15]. Besides, reaction temper-
consequently display (2) higher temperature rise and (3) higher
ature rise (T) was also positively correlated to polymerization
flexural modulus, flexural strength and hardness, compared to
shrinkage [16].
1282 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289
a conventional CQ-based experimental composite cured for 20
or 40 s at 1000 mW/cm2 .
2. Materials and methods
2.1. Materials
Two experimental composites materials were prepared, based
on a 70/30 wt% Bis-GMA/TegDMA resin (Sigma–Aldrich),
using two different photoinitiators systems in equimolar
concentration (1.34 × 10−5 mol cm−3 ): either camphorquinone
and dimethylaminoethyl methacrylate (altogether CQ, from
Sigma Aldrich) or Lucirin-TPO (TPO, from BASF). Initia-
tors accounted for 0.2/0.8 wt% and 0.42 wt% for CQ and
TPO, respectively. Fillers were silanated barium glass (G018-
186/K6, d50 = 3 ± 1 ␮m, Schott AG, Landshut Germany) and
methacrylsilane treated fumed silica (12 nm, Evonik Indus-
tries, Germany) introduced in amounts of 65/10 wt% respec-
tively. The fillers were dispersed using a dual asymmetric
centrifuge (Speed mixer, FlackTek, USA).
2.2. Flexural bending test
Rectangular specimens for the flexural bending test were
polymerized in Teflon moulds of dimensions 25 × 2 × 2 mm
(n = 5 following ISO4049). Light-curing was carried out with an
AURA light device (Lumencor, USA) providing narrow spec-
tral outputs. Irradiation could be applied independently at
two wavelength ranges: 395–415 nm and 455–485 nm for TPO
and CQ-based composites, respectively (Fig. 1). Light irradi-
ance calibrations were carried out as described previously [17].
Irradiation was applied in 5 successive and non-overlapping
cycles starting at the center of the sample, then extended
towards the ends. During each irradiation, Mylar films were
used to cover the top and bottom surfaces of the samples and
opaque masks with a 5 mm diameter window (the dimension
Fig. 1 – Emission spectra of the AURA light engine two
independent outputs (Lumencor) compared to the
absorption spectra of Camphorquinone (CQ) and
Lucirin-TPO (TPO).
of the light guide) were used to cover the adjacent sample
areas and thereby reduce overlapping irradiation areas due
to light diffusion. The bars were stored for 24 h in the dark at
room temperature before being tested.
The CQ-based composites (controls) were irradiated for 20
and 40 s at 1000 mW/cm2 . TPO-composites were irradiated at
three different irradiances for: 2000 mW/cm2 for 9, 3, 1 or 0.5 s,
1000 mW/cm2 for 40 s 20, 9, 3, 1 or 0.5 s and 500 mW/cm2 for 9,
3, 1 or 0.5 s.
Flexural testing was conducted using a universal testing
machine (Lloyd LRX Plus, Lloyd Instruments), with a 20 mm
span between supports and 0.75 mm/min crosshead speed.
Flexural modulus was calculated according to Eq. (1) and flex-
ural strength was calculated according to Eq. (2), where F is
the load (N), l is the distance between supports (mm), w and t
are the width and thickness of the sample (mm) and d is the
deflection due to the load F (mm).
Fl3
E= (1)
4wt3 d
3Fl
rupt = (2)
2wt2
2.3. Vickers hardness
5 mm diameter, 2 mm thickness disc-shaped specimens were
prepared in split Teflon moulds and were irradiated follow-
ing the same protocol as described in the previous section
(n = 3). Following irradiation, the Mylar films were removed,
samples were extracted from the moulds and were stored
for 24 h in the dark at room temperature before being tested.
A first series of hardness measurements was then carried
out. Subsequently, samples were incubated for another 24 h
in 75/25 vol% ethanol/H2 O; period after which hardness was
again measured. In both cases, measurements were carried
out on the upper surface (two per surface) with a microhard-
ness tester (Durimet,Leitz, Wetzlar, Germany) using a 100 g
load (F = 0.1 kgf) and 30 s indentation period. Hardness calcu-
lation was done using mean diagonal values of indentations
(d) and Eq. (3).
1.854F
HV = (3)
d2
2.4. Polymerization stress
The stress generated during the polymerization reaction was
measured using a single cantilever device, the Bioman instru-
ment, as previously described [26]. In the present study the
specimen layer was restricted to 0.9 ± 0.05 mm to limit the
flowing of the material. The light source was located below
the glass slide at a distance of 7.5 mm from the bottom
surface of the sample and calibrations were repeated in order
to obtain actual irradiances of 500, 1000 and 2000 mW/cm2 .
The force (stress) generated by the material was measured via
the strain-gage signal of the cantilever load cell and using Eq.
(4) (n = 3).
kV
p = (4)
S
 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289
where V is the voltage difference read at the sensor during
the measurement, k is a constant such as k V is equivalent
to Newtons and S is the surface of the sensor (mm2 ).
Maximum polymerization stress was determined as the
value reached after 4 min of acquisition and rates were cal-
culated as the first derivative against time.
2.5. Temperature analysis
The temperature variations associated with polymerization
and irradiation were monitored using an experimental setup.
A composite specimen was placed in a Teflon mold (2 mm
thick, 5 mm inner diameter), itself resting on a molar tooth
which was machined to leave a 0.6 mm thick dentin roof,
accommodate a temperature sensor in the pulp chamber
and present a flat occlusal surface (Fig. 2(a)). The tooth was
partially immersed in a temperature-controlled water bath
(37 ± 1 ◦ C) as shown in Fig. 2 b). Irradiation was carried out
as in other analyses, with the light guide in contact with the
Mylar strip covering the composite specimen. Temperature
was recorded with a 4-wire platinum sensor and data acqui-
sition module (HSRTD-3-100-A-1 M and PT-104A respectively,
Omega, UK) every 0.1 s with an accuracy of 0.15 ◦ C (n = 3). The
maximum temperature changes are reported (T).
2.6. Degree of polymerization
Real-time DC measurement was performed using Fourier
Transform Near Infrared Spectroscopy (FT-NIRS) as previously
described (n = 3) [19]. Briefly, resin composites were placed in
white Teflon disk-shaped moulds (12 × 2 mm) with lower sur-
faces in contact with a glass microscope slides (1.2 mm thick).
Degrees of conversion were determined in real-time by moni-
toring the changes in peak height associated with the reaction
of vinyl CH2 . DCs reported in Table 1 represent mean values at
110 s of monitoring.
Fig. 2 – Tooth preparation (a), general setup for pulpal
temperature rise analysis (b) and evolution of temperature
for TPO and CQ composites irradiated at 1000 mW/cm2 (c).
1283
2.7. Microleakage
15 wisdom teeth extracted for therapeutic reasons were dis-
tributed into 3 groups (n = 5), CQ 20 s, TPO 1 s and TPO 3 s,
1000 mW/cm2 . They were stored in 0.5% chloramine solution
at 6 ◦ C until cavity preparation. After been cleaned (soft tissue
and calculus were removed), two class V cavities (2 mm depth,
4 mm width and 2 mm height) were prepared per tooth on the
buccal and lingual surfaces under continuous water irrigation.
The cavities extended 1 mm on either side of the cement-
enamel junction, with a bevel on the enamel side. The gold
standard total-etch adhesive Optibond FL (Kerr) was selected
and applied following the manufacturer’s instructions. Briefly,
the cavity was conditioned with 37.5% phosphoric acid (Gell
Etchant, Kerr) during 30 s for enamel and 15 s for dentin. The
primer was applied, followed by the bonding, and light irradi-
ation was carried out using the AURA light engine (emitting
at 455–485 nm) during 20 s at 1000 mW/cm2 to ensure proto-
col consistency. Cavities were then restored in a single layer
using the TPO or CQ-based resin composites and using the
appropriate lamp outputs as described in previous sections.
Restored teeth were then polished using Sof-Lex Finishing
discs (3M-ESPE) and root apices were sealed with a resin-
modified glass-ionomer (Fuji II, GC) to prevent the dye from
penetrating through the root canals. The teeth were then
thermocycled for 100 cycles between two water-baths kept at
temperatures of 1.4 ± 0.5 ◦ C and 60 ± 1 ◦ C. A cycle consisted of
10 s immersion in one bath, 30 s transfer-time and another 10 s
immersion in the other bath. Immediately following thermo-
cycling the teeth were placed in a 0.5% basic fushin solution for
24 h and kept at 37 ◦ C ± 1 ◦ C. Finally, the teeth were embedded
in epoxy resin and cut to obtain three 0.5 mm-thick sections
per tooth. Given the inherent structural inhomogeneity of
dentin, each section was considered as an independent sys-
tem, yielding a total of 30 sections per group to be scored (3
sections per cavity, 2 cavities per tooth and 5 teeth per group).
Microleakage was recorded under 10x magnification at the
enamel and dentin interface by a blinded evaluator: a score
of ‘0 was assigned in the absence of dye penetration, while
‘1 or ‘2 were assigned for dye penetration along the enamel
or gingival floor up to or exceeding half the restoration depth,
respectively. A score of ‘3 was given when the dye stained the
axial wall.
2.8. Statistical analysis
Statistical analyses were performed using the JMP statisti-
cal software (SAS Institute Inc.). All data sets were analyzed
using one-way analysis of variance (ANOVA) and with Tukey
tests (˛ = 0.05) except for the microleakage results. In this case,
paired group comparison were carried out using the non-
parametric Wilcoxon test (˛ = 0.05).
3. Results
Polymerization occurred faster in TPO-composites compared
to the controls as seen in Fig. 3. Final DCs were significantly
higher (p < 0.05) in TPO-composites compared to CQ controls
for irradiation times equal to or greater than 3 s (Table 1).
1284 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289

Table 1 – Mean values for final DC, maximum polymerization stress, stress rate and Tmax of all TPO and
CQ-composites. Standard deviation in parentheses. Same superscript connects statistically similar values (p < 0.05).
Irradiation time (s)/Irradiance Radiant energy Final DC (%) Final  p (MPa) Stress rate T (◦ C)
(mW/cm2 ) (J/cm2 ) (MPa/s)
TPO
0.5/500 0.25 42 (2)H 2.4 (0.2)F 0.9 (0.1)E 0.43 (0.05)I
1/500 0.5 56 (1)F,G 2.9 (0.2)E,F 1.3 (0.1)D,E 0.54 (0.05)H,I
3/500 1.5 65 (1)B,C 3.4 (0.2)D,E 1.2 (0.1)D,E 0.75 (0.03)E,F,G
9/500 4.5 68 (1)A,B 4.4 (0.3)A,B 1.3 (0.1)D,E 0.92 (0.01)D,E
0.5/1000 0.5 54 (2)G 3.2 (0.2)D,E 1.6 (0.3)C,D 0.59 (0.03)G,H,I
1/1000 1 63 (2)D,E 3.7 (0.2)B,C,D 1.9 (0.2)A,B,C 0.73 (0.06)F,G
3/1000 3 70 (1)A 4.3 (0.2)A,B,C 1.9 (0.1)A,B,C 0.83 (0.03)E,F
9/1000 9 69 (1)A 4.6 (0.1)A 1.7 (0.1)B,C,D 1.14 (0.09)C
20/1000 20 70 (1)A 4.6 (0.3)A 2.0 (0.3)A,B,C 1.52 (0.04)B
40/1000 40 69 (1)A 4.9 (0.1)A 2.0 (0.1)A,B,C 1.88 (0.06)A
0.5/2000 1 64 (1)C,D 3.6 (0.2)C,D 2.3 (0.4)A,B 0.68 (0.03)F,G,H
1/2000 2 68 (2)A,B 3.5 (0.4)D,E 2.3 (0.2)A,B 0.81 (0.07)E,F
3/2000 6 70 (1)A 4.6 (0.3)A 2.1 (0.2)A,B,C 1.02 (0.13)C,D
9/2000 18 70 (1)A 4.9 (0.3)A 2.5 (0.2)A 1.45 (0.05)B
CQ
20/1000 20 57 (1)E,F 3.5 (0.2)D,E 0.2 (0.1)F 1.21 (0.08)C
40/1000 40 60 (1)E,F 3.5 (0.3)D,E 0.2 (0.1)F 2.05 (0.07)A

Under optimal conditions, TPO-composites display an average
10% conversion increase. Polymerization stresses increased in
a logarithmic pattern with time, maximum stress rate being
reached in less than 3 s for TPO-composites and between 5 and
10 s for CQ-controls (Fig. 4). T increased with time, reaching
a maximum between 20 and 40 s, depending on the curing
parameters (Fig. 2(c)). In the case of CQ-controls, a sharp
drop of T curves could be observed, while T decrease was
more progressive for all TPO-composites. For 20 and 40 s cur-
ing modes, Tmax was reached when the curing light was
switched off (as indicated by red arrows on Fig. 2 c)). However,
Tmax was reached 10–20 s after the light was switched off
for TPO-materials cured during 9 s or less.
When plotting maximum stress and Tmax against
radiant energy, different trends were observed (Fig. 5 c)).
In TPO-composites, the maximum polymerization stress
reached a plateau at approximately 5 J/cm2 while Tmax
continued to increase linearly above that level of radiant
exposure. With similar irradiation parameters (20 or 40 s at
1000 mW/cm2 ), TPO-composites and controls exhibited simi-
lar Tmax as controls, but the latter showed lower maximum
stress (Table 1, Fig. 5). No improvements were observed when
curing time was increased from 20 to 40 s for CQ controls:
they displayed similar DCmax and maximum polymeriza-
tion stress. TPO-composites displayed higher polymerization
stresses compared to the CQ controls when irradiated for 3 s
or more for irradiances greater than 1000 mW/cm2 . Maximum
stress rates were all significantly higher for TPO-composites,
by at least a factor 5 (Fig. 3 and Table 1).
Flexural modulus was in most cases significantly higher
in TPO-composites compared with controls, except for three
curing parameters (0.5 and 1 s at 500 mW/cm2 and 0.5 s at
1000 mW/cm2 (Fig. 6(a)). Flexural strength values were simi-
lar between TPO and controls except for the 500–0.5 protocol
(Fig. 6 b)). Upper surface hardness increased both with irradia-
tion time and irradiance in TPO-composites. Hardness values
were significantly higher than in controls for irradiation times
greater than 1s at 1000 and 2000 mW/cm2 (Fig. 6(c). After incu-
bation in the ethanol solution, hardness dropped by 39 to 47%
in controls while the drop in TPO-composites ranged between
21 and 36%. Hardness values of TPO-composites after storage
in ethanol solution were all similar to or higher than controls.
No microleakage was observed at the enamel/composite
interface for all samples. For the dentin/composite interface,
wilcoxon’s test on paired groups indicated a significantly
higher microleakage for the TPO 1 s group compared to CQ
20 s (p = 0.0274) and TPO 3 s (p = 0.0102). TPO 3 s and CQ 20 s did
not show any significant difference (p = 0.6894). Fig. 7 shows
the fraction of scores attributed to the different groups.
4. Discussion
The three null hypotheses, namely that TPO-based com-
posites cured at short curing times and displaying high DC
would result in (1) higher polymerization stresses, (2) higher
temperature rises, and (3) higher flexural modulus, flexural
strength and hardness compared to a conventional CQ-based
composite, were all rejected. First, TPO-based materials gen-
erated higher polymerization stresses than those generated
by the controls, but only for some curing protocols. Second,
temperature increases were in most cases lower than those
measured for the CQ controls. Third, flexural modulus and
hardness were significantly higher in TPO-composites than
in CQ-controls for most but not all curing protocols, while
flexural strength of TPO-composites and CQ-controls were
similar (except for one condition).
The curing kinetics measured with FT-NIRS were in
accordance with previous literature comparing CQ and TPO
materials [19], where TPO-based composites reached their
maximum DC in less than 5 s when using high irradiances
(1000 mW/cm2 or above). In the present work, 95% of DC
was reached in about 2–3 s (Fig. 3). In accordance with our
previous study on the same materials/curing conditions [17],
TPO-composites did not benefit from an increase in irradi-
ance and/or irradiation time above a given threshold. More
precisely this corresponded to irradiation protocols of 3 s
irradiation time at 1000 and 2000 mW/cm2 (3–6 J/cm2 ) and 9 s
 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289
Fig. 3 – Evolution of DCs over time in TPO-composites
irradiated at 2000 mW/cm2 (a), TPO and CQ-composites
irradiated at 1000 mW/cm2 (b) and TPO-composites
irradiated at 500 mW/cm2 (c). Red vertical arrows indicate
irradiation times and are reproduced with the same length
on (a), (b) and (c).
at 500 mW/cm2 (4.5 J/cm2 ) (Table 1). These findings confirmed
what was previously reported using electron paramagnetic
resonance [17], wherein plotting radical concentration against
radiant energy revealed a plateau following 3 J/cm2 , and there-
fore, 3 s of irradiation at 1000 mW/cm2 yielded optimal curing
with sub-optimal states below that threshold. In the present
work, the existence of a radiant exposure threshold was
confirmed for polymerization stress, with a plateau value
reached around 3–5 J/cm2 (Fig. 5(b)).
In comparison, the build-up of polymerization stress in CQ-
based controls was much slower than TPO-composites, with
no difference between 20 or 40 J/cm2 . This is in accordance
with the DC data (Fig. 3), where DC of controls reached 95% of
the final DC value after 20 s of irradiation. This slower stress
1285
Fig. 4 – Evolution of polymerization stress for
TPO-composites (black) and CQ-composites (grey)
irradiated at 1000 mW/cm2 (a). Evolution of stress rates for
the same materials (b).
build-up could be expected, since the polymerization reac-
tion of CQ-based composites has been reported to depend
more on irradiation time rather than on irradiance [8,27,28].
Specifically regarding the direct comparison of CQ vs TPO
materials, it was shown that the latter cure better when irra-
diated for 45 s at 400 mW/cm2 than at 6 s at 3000 mW/cm2
[19] (both 18 J/cm2 ). Altogether these results tend to indi-
cate that while CQ-based RBC require long irradiation times,
TPO-composites have the ability to cure in a fraction of that
time, hence requiring a much lower energy to reach full cure
[17].
Polymerization stress is a corollary of chain lengthen-
ing, entanglement and cross-linking during curing [10]. The
evolution of stress in resin composites occurs non-linearly
with conversion [12,29]. In dimethacrylate resins or resin
composite systems, polymerization proceeds following a
diffusion-controlled mechanism [30] and in final stages, cross-
linking induces a major change in viscosity. Beyond this
point, any increase in conversion results in further contrac-
tion in a highly rigid material and hence increased stress
and the present results for TPO-composites follow such a
pattern (Table 1). As noted previously, at radiant energies
of 3 J/cm2 and above, the stress reaches a plateau with
final values of about 4–5 MPa (Fig. 5 b)). Since maximum DC
is reached at a similar threshold, the increase, then satu-
ration of maximum stress with increasing radiant energy
1286 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289
Fig. 5 – Average final DC versus maximum polymerization
stress (a), maximum polymerization stress and Tmax
versus radiant energy (b) and (c) respectively for all TPO
(blue) and CQ-composites (white).
is logically linked to the evolution in the final stages of
polymerization.
Contrary to stress curves, Tmax curves continued
increasing in a linear fashion above the 3 J/cm2 radiant
energy threshold, and up to 40 J/cm2 (Fig. 5(c)). To discrimi-
nate between T due to the reaction exotherm and T from
the curing light, it could be approximated that the tem-
perature rise purely related to the polymerization reaction
corresponded to T in specimens with saturated DC and
short curing time (e.g. 3 s at 1000 mW/cm2 ). For these param-
eters, the light was indeed switched off long before T curves
reached their maximum (Fig. 2(c)). The amount of heat purely
due to the chemical reaction was then about 0.8 ◦ C for TPO-
composites (through 0.6 mm dentin, and with the specific
experimental setup of this work). Any additional temper-
ature increase should then be attributed to excess energy
provided by the LCU and not used by the composite for addi-
tional propagation [13,16,19]. Following this reasoning, the
increase of T at short curing times is concomitant to the
increase of stress and conversion until the threshold. Above
the latter however, longer irradiation times induced compos-
ite over-heating without any noticeable change in conversion
measured within the timeframe of the experiment. Simi-
larly, for CQ-composites, the increase from 20 to 40 s did not
result in DC increase, but an almost two-fold T increase was
observed. As mentioned in previous work, curing parameters
should be optimized to reach optimal curing but at the same
time avoid unnecessary composite – and pulp – overheating
[31].
It has been previously described that the onset of vitri-
fication is delayed in TPO-composites [19,20] compared to
CQ-based materials, possibly as a result of transfer reactions
due to the TPO system [32]. In methacrylate networks, such
mechanism is known to promote higher conversion degrees
and reduced stress [33]. This trend was confirmed in the
present study, since similar stresses were indeed observed in
CQ- and TPO-based materials, despite the higher DCs mea-
sured in the latter.
Vitrification is the stage where the polymer matrix fully
rigidifies. It may be defined as the point where kinetically,
the exotherm of the reaction itself is no longer sufficient to
provide mobility to propagating radicals in an increasingly vis-
cous medium. In this situation any additional heat provided by
the LCU to the system could be expected to postpone that vitri-
fication by providing additional mobility. In the present study,
we used an open system where the temperature measurement
had high clinical relevance but lacked the adiabatic quality of
systems such as differential scanning calorimetry. However,
given the precision of our measurement, it still made sense
comparing temperature increases in relation with matrix for-
mation. The additional heat provided by the LCU was clearly
not enough to significantly increase conversion, possibly indi-
cating that T increase was insufficient to increase Tg in
TPO-composites [34]. Here Tg is defined as the temperature
above which a cross-linked matrix fully relaxes [35]. In order
for the heating from the LCU to have a major impact on DC, it
should preferably occur during polymerization, i.e. when T
of both LCU and reaction are cumulated. This would be diffi-
cult to observe in TPO-composites since the polymer matrix
was already highly crosslinked after 3 s as indicated by FT-
NIRS (Fig. 2) and previous results [19,20]. In CQ-composites
where polymerization is inherently slower, this is more likely
to occur, but again it did not appear to be the case from DC
values, which are at best similar, and in general lower than
TPO-composites. Hence, these observations tended to show
that the levels of temperature rise in the materials were insuf-
ficient to significantly move Tg , but also relatively safe for
the pulp tissues when placed on a 0.6 mm thick wet dentin
layer (<5.5 ◦ C [14]). Again, the measurement of temperature
rise in the pulp chamber was probably very dependent on the
experimental setup, and it is very difficult to compare tem-
perature between studies. This could account for the smaller
s values reported here as compared to other in-vitro works
[16,36].
 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289 1287

Shrinkage stress, is known to build-up quickly as a result

of monomer to polymer conversion, but then increases at a
slower rate after the curing light is off, as a result of “post-cure
conversion or physical densification” [29]. In case of TPO-
composites, larger amounts of free volume are likely to be
generated during polymerization as faster rates of reaction
in BisGMA/TegDMA systems yield greater excess free volume
in general [37]. The “ultra-fast” kinetics of TPO-composites
led to high stress rates compared to the CQ-controls. The
very fast development of stress may be detrimental to the
integrity of the adhesive interface. However the present
microleakage results along with that of a previous study [20]
do not support such claim. It is possible that higher DC in
TPO-composites counterbalances the impact of high stress
levels and/or rates by providing a more robust and stable
interface.
Finally, it should be pointed out that in works study-
ing the evolution of strain (which induces stress for a
constrained material) with the polymerization exotherm in
Fig. 7 – Distribution of microleakage scores for class V
CQ-based materials, a change in slope was observed precisely
cavities restored with experimental TPO or CQ-based resin
when light was switched off [16,38]. This latter phenomenon
composites (TPO 1 s, TPO 3 s or CQ 20 s).
was assigned to a thermal contraction, a thermodynamic
requirement caused by the sudden interruption of energy sup-
plied by the LCU and a return to room temperature. In Fig. 2(c),
a similar observation could be made for CQ-composites, with contraction due to thermal effects, or shrinkage resulting from
a sharp temperature drop at 20 or 40 s, i.e. exactly when the the polymerization [13]. Also, it has been shown that the heat
light is switched off. However, such a drop was not observed capacity (Cp ) of a polymer is lower than that of the equiva-
in TPO-composites, at any irradiation time. The reasons for lent monomer [16], which may indicate to some extent that a
this difference remain unclear, since the exotherm curve can polymer is more conductive than its monomer. The lack of
be influenced by several competing events, i.e. expansion or significant temperature drop at shut-off (even at 20 or 40 s

Fig. 6 – Flexural modulus (GPa) (a), Flexural strength (MPa) (b), Hardness (HV) (c) and hardness after 24 h in 75 vol% EtOH/H2 O
(MPa) (d) at each irradiation time (s) for TPO-composites irradiated at 500, 1000 and 2000 mW/cm2 and TPO and
CQ-composites irradiated at 1000 mW/cm2 . Bars connected with the same letter have statistically similar values (p < 0.05).
Error bars indicate standard deviations.
1288 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289
irradiation times) for TPO-composites as compared to CQ-
controls highlights some structural differences, accounting for
the fact that although both composites are vitrified, there is
a mechanism in TPO-composites compensating for thermal
contraction, i.e. generating some heat.
General improvement of mechanical properties
The general increase in mechanical properties observed for
TPO-materials for most curing conditions compared to con-
trols is likely explained by the higher DC values measured with
TPO materials. The increase in DC has indeed been shown to
result in an increase in modulus [39] and hardness [39,40],
due to a decrease in un-polymerized content, acting as a
plasticizer, and to a higher degree of cross-linking. This is
in line with previous work investigating the same TPO and
CQ-based model composites, where better cured materials
were shown to release less monomers and hence to contain a
smaller fraction of un-polymerized content [17]. In the present
work, higher DCs in TPO-composites having reached the DC
plateau presented more rigid materials: the average flex-
ural modulus was 8.6 ± 0.6 and 6.4 ± 0.4 GPa for TPO 1000–3 s
and CQ-composites 1000–20 s respectively. On the contrary,
no significant differences were observed for flexural strength
(except for the undercured condition), with an average value of
91.5 MPa. Higher surface hardnesses were also observed under
most irradiation protocols in TPO-composites compared to
the control (53.7 ± 0.7 HV compared to 49.4 ± 0.7 HV for TPO
1000–3 s and CQ-composites 1000–20 s respectively). Hardness
is frequently used as an indirect measurement of DC, but is
also an indicator of cross-linking [9,41]. Moreover, a more pro-
nounced drop in hardness was observed in CQ-controls after
incubation in the ethanol solution. This could be explained by
a higher sorption, inherent to a less converted matrix. Inter-
estingly and contrary to that reported for DC, shrinkage stress
increased and flexural modulus and upper surface hardness
were improved in TPO-composites with the increase of irradi-
ation time beyond 3 s (Fig. 6). This was surprising, especially
since the flexural modulus (significantly) and strength (not
significantly) decreased for irradiation times longer than 3 s
at 1000 mW/cm2 . The improvement of hardness could not be
explained.
The values obtained for the flexural modulus, flexural
strength and polymerization stress were consistent with
those reported in previous work, that studied 50/50 wt% Bis-
GMA/TegDMA resin composites filled at 55 vol% and using the
same concentrations of photoinitiators [20].
5. Conclusion
TPO-based resin composites irradiated with a curing lamp
adapted to the absorption spectrum of Lucirin-TPO presented
fast curing kinetics and improved DC compared to CQ-based
formulations. Complementing the results of a previous study,
the present work demonstrates that TPO-composites irradi-
ated for very short curing times (1 s at 1000 mW/cm2 ) not only
display reduced monomer release, but improved mechan-
ical properties versus CQ-composites with an increase in
25% for flexural modulus (8.1 GPa compared to 6.4 GPa) and
40% for hardness measured after EtOH/H2O sorption (41.4 HV
compared to 26.1 HV). For such curing protocol the polymer-
ization stress generated in TPO-composites was similar to
the materials incorporating CQ but the stress rate was much
higher. However microleakage was similar between CQ and
TPO-composites irradiated for 3 s, suggesting that increased
conversion may counterbalance higher stress and stress rate
in interfacial cohesion. Due to shorter irradiation times, the
total exotherm in TPO-composites was significantly lower.
These results further support the use of Lucirin-TPO as a
replacement of CQ in commercial products using an appro-
priate curing wavelength.
Acknowledgements
Julian Leprince is a Postdoctoral Researcher at F.R.S.-FNRS. The
authors are very grateful to Matthieu Gilli, Chloe Hardy and
Thibaut Hollaert for their precious help with the microleak-
age experiments, and to the department of Stomatology and
Maxillo-facial Surgery for supplying the teeth. The authors are
also grateful to BASF for supplying the Lucirin-TPO.
references
[1] Hickel R, Manhart J. Longevity of restorations in posterior
teeth and reasons for failure. J Adhes Dent 2001;3:45–64.
[2] Van Nieuwenhuysen JP, D’Hoore W, Carvalho J, Qvist V.
Long-term evaluation of extensive restorations in
permanent teeth. J Dent 2003;31:395–405.
[3] Opdam NJ, Bronkhorst EM, Roeters JM, Loomans BA. A
retrospective clinical study on longevity of posterior
composite and amalgam restorations. Dent Mater
2007;23:2–8.
[4] Opdam NJ, Bronkhorst EM, Loomans BA, Huysmans MC.
12-year survival of composite vs. amalgam restorations. J
Dent Res 2010;89:1063–7.
[5] Gutiérrez-Salazar MdP, Reyes-Gasga J. Microhardness and
chemical composition of human tooth. Mat Res
2003;6:367–73.
[6] Kinney JH, Marshall SJ, Marshall GW. The mechanical
properties of human dentin: a critical review and
re-evaluation of the dental literature. Crit Rev Oral Biol Med
2003;14:13–29.
[7] Jun S-K, Kim D-A, Goo H-J, Lee H-H. Investigation of the
correlation between the different mechanical properties of
resin composites. Dent Mater J 2013;32:48–57.
[8] Dewaele M, Asmussen E, Peutzfeldt A, Munksgaard EC,
Benetti AR, Finne G, Leloup G, Devaux J. Influence of curing
protocol on selected properties of light-curing polymers:
degree of conversion, volume contraction, elastic modulus,
and glass transition temperature. Dent Mater
2009;25:1576–84.
[9] Ferracane JL. Correlation between hardness and degree of
conversion during the setting reaction of unfilled dental
restorative resins. Dent Mater 1985;1:11–4.
[10] Ferracane JL. Developing a more complete understanding of
stresses produced in dental composites during
polymerization. Dent Mater 2005;21:36–42.
[11] Braga RR, Ballester RY, Ferracane JL. Factors involved in the
development of polymerization shrinkage stress in
resin-composites: a systematic review. Dent Mater
2005;21:962–70.
 d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 1280–1289
[12] Lu H, Stansbury JW, Bowman CN. Towards the elucidation of
shrinkage stress development and relaxation in dental
composites. Dent Mater 2004;20:979–86.
[13] Mucci V, Arenas G, Duchowicz R, Cook WD, Vallo C.
Influence of thermal expansion on shrinkage during
photopolymerization of dental resins based on
bis-GMA/TEGDMA. Dent Mater 2009;25:103–14.
[14] Zach L, Cohen G. Pulp response to externally applied heat.
Oral Surg Oral Med Oral Pathol 1965;19:515–30.
[15] Baldissara P, Catapano S, Scotti R. Clinical and histological
evaluation of thermal injury thresholds in human teeth: a
preliminary study. J Oral Rehabil 1997;24:791–801.
[16] Alnazzawi A, Watts DC. Simultaneous determination of
polymerization shrinkage, exotherm and thermal expansion
coefficient for dental resin-composites. Dent Mater
2012;28:1240–9.
[17] Randolph LD, Palin WM, Bebelman S, Devaux J, Gallez B,
Leloup G, Leprince JG. Ultra-fast light-curing resin composite
with increased conversion and reduced monomer elution.
Dent Mater 2014;30:594–604.
[18] Moszner N, Fischer UK, Ganster B, Liska R, Rheinberger V.
Benzoyl germanium derivatives as novel visible light
photoinitiators for dental materials. Dent Mater
2008;24:901–7.
[19] Leprince JG, Hadis M, Shortall AC, Ferracane JL, Devaux J,
Leloup G, Palin WM. Photoinitiator type and applicability of
exposure reciprocity law in filled and unfilled photoactive
resins. Dent Mater 2011;27:157–64.
[20] Palin WM, Hadis MA, Leprince JG, Leloup G, Boland L,
Fleming GJP, Krastl G, Watts DC. Reduced polymerization
stress of MAPO-containing resin composites with increased
curing speed, degree of conversion and mechanical
properties. Dent Mater 2014;30:507–16.
[21] Santini A, Miletic V, Swift MD, Bradley M. Degree of
conversion and microhardness of TPO-containing
resin-based composites cured by polywave and monowave
LED units. J Dent 2012;40:577–84.
[22] Pongprueksa P, Miletic V, Janssens H, Van Landuyt KL, De
Munck J, Godderis L, Van Meerbeek B. Degree of conversion
and monomer elution of CQ/amine and TPO adhesives. Dent
Mater 2014.
[23] Hadis M, Leprince JG, Shortall AC, Devaux J, Leloup G, Palin
WM. High irradiance curing and anomalies of exposure
reciprocity law in resin-based materials. J Dent
2011;39:549–57.
[24] Jandt KD, Mills RW. A brief history of LED
photopolymerization. Dent Mater 2013;29:605–17.
[25] Leprince JG, Palin WM, Hadis MA, Devaux J, Leloup G.
Progress in dimethacrylate-based dental composite
technology and curing efficiency. Dent Mater 2013;29:
139–56.
[26] Watts DC, Marouf AS, Al-Hindi AM. Photo-polymerization
shrinkage-stress kinetics in resin-composites: methods
development. Dent Mater 2003;19:1–11.
1289
[27] Feng L, Suh BI. Exposure reciprocity law in
photopolymerization of multi-functional acrylates and
methacrylates. Macromol Chem Physic 2007;208:295–306.
[28] Leprince JG, Lamblin G, Devaux J, Dewaele M, Mestdagh M,
Palin WM, Gallez B, Leloup G. Irradiation modes’ impact on
radical entrapment in photoactive resins. J Dent Res
2010;89:1494–8.
[29] Stansbury JW. Dimethacrylate network formation and
polymer property evolution as determined by the selection
of monomers and curing conditions. Dent Mater
2012;28:13–22.
[30] Anseth KS, Newman SM, Bowman CN. Polymeric dental
composites: Properties and reaction behavior of
multimethacrylate dental restorations. In: Peppas N, Langer
R, editors. Biopolymers II. Springer Berlin Heidelberg; 1995.
p. 177–217.
[31] Leprince J, Devaux J, Mullier T, Vreven J, Leloup G.
Pulpal-temperature rise and polymerization efficiency of
LED curing lights. Oper Dent 2010;35:220–30.
[32] Neumann MG, Schmitt CC, Ferreira GC, Corrêa IC. The
initiating radical yields and the efficiency of polymerization
for various dental photoinitiators excited by different light
curing units. Dent Mater 2006;22:576–84.
[33] Pfeifer CS, Wilson ND, Shelton ZR, Stansbury JW. Delayed
gelation through chain-transfer reactions: mechanism for
stress reduction In methacrylate networks. Polymer
2011;52:3295–303.
[34] Lovell LG, Lu H, Elliott JE, Stansbury JW, Bowman CN. The
effect of cure rate on the mechanical properties of dental
resins. Dent Mater 2001;17:504–11.
[35] Stutz H, Illers KH, Mertes J. A generalized theory for the glass
transition temperature of crosslinked and uncrosslinked
polymers. J Polym Sci Pol Phys 1990;28:1483–98.
[36] Hannig M, Bott B. In-vitro pulp chamber temperature rise
during composite resin polymerization with various
light-curing sources. Dent Mater 1999;15:275–81.
[37] Pfeifer CS, Shelton ZR, Braga RR, Windmoller D, Machado JC,
Stansbury JW. Characterization of dimethacrylate polymeric
networks: a study of the crosslinked structure formed by
monomers used in dental composites. Eur Polym J
2011;47:162–70.
[38] Atai M, Watts DC, Atai Z. Shrinkage strain-rates of dental
resin-monomer and composite systems. Biomaterials
2005;26:5015–20.
[39] Ferracane JL, Greener EH. The effect of resin formulation on
the degree of conversion and mechanical properties of
dental restorative resins. J Biomed Mater Res 1986;20:121–31.
[40] Leprince JG, Leveque P, Nysten B, Gallez B, Devaux J, Leloup
G. New insight into the depth of cure of
dimethacrylate-based dental composites. Dent Mater
2012;28:512–20.
[41] Watts DC, Amer OM, Combe EC. Surface hardness
development in light-cured composites. Dent Mater
1987;3:265–9.