Stress relaxation via addition-fragmentation chain transfer in high T g, high conversion methacrylate-based systems

Article abstract of DOI:10.1021/ma300228z

To reduce shrinkage stress which arises during the polymerization of cross-linked polymers, allyl sulfide functional groups were incorporated into methacrylate polymerizations to determine their effect on stress relaxation via addition-fragmentation chain transfer (AFCT). Additionally, stoichiometrically balanced thiol and allyl sulfide-containing norbornene monomers were incorporated into the methacrylate resin to maximize the overall functional group conversion and promote AFCT while also enhancing the polymer's mechanical properties. Shrinkage stress and reaction kinetics for each of the various functional groups were measured by tensometry and Fourier-transform infrared (FTIR) spectroscopy, respectively. The glass transition temperature (T _g) and elastic moduli (E′) were measured using dynamic mechanical analysis. When the allyl sulfide functional group was incorporated into dimethacrylates, the polymerization-induced shrinkage stress was not relieved as compared with analogous propyl sulfide-containing resins. These analogous propyl sulfide-containing monomers are incapable of undergoing AFCT while having similar chemical structure and cross-link density to the allyl sulfide-containing methacrylates. Here, a monomethacrylate monomer that also contains a cyclic allyl sulfide (PAS) was found to increase the cross-linking density nearly 20 times as compared to an analogous monomethacrylate in which the allyl sulfide was replaced with an ethyl sulfide. Despite the much higher cross-link density, the PAS formulation exhibited no concomitant increase in stress. Thiol-norbornene resins were copolymerized in PAS to promote AFCT as well as to synergistically combine the ring-opening benefits associated with the thiol-ene reaction. AFCT resulted in a 63% reduction of polymerization stress and a 45 °C enhancement of the glass transition temperature in the allyl sulfide-containing thiol-norbornene-methacrylate system compared with rubbery dimethacrylates. When compared with conventional glassy dimethacrylates, this combined system has less than 10% of the typical shrinkage stress level while having similarly excellent mechanical properties.

Full text of DOI:10.1021/ma300228z

Article
pubs.acs.org/Macromolecules
Stress Relaxation via Addition−Fragmentation Chain Transfer in
High T , High Conversion Methacrylate-Based Systems
g
†
‡
§
§
Hee Young Park, Christopher J. Kloxin, Ahmed S. Abuelyaman, Joe D. Oxman,
and Christopher N. Bowman*
,
†,∥,⊥
†Department of Chemical and Biological Engineering, Jennie Smolly Caruthers Biotechnology Building, University of Colorado at
Boulder, Boulder, Colorado 80309-0596, United States
‡
Department of Materials Science and Engineering and Department of Chemical Engineering, University of Delaware, Newark,
Delaware 19716, United States
§
3
M ESPE Dental Products, 3M Center, 260-2A-10, St. Paul, Minnesota 55144-1000, United States
∥
Department of Restorative Dentistry, School of Dentistry, University of Colorado at Denver, Aurora, Colorado 80045, United States
⊥BioFrontiers Institute, University of Colorado at Boulder, Boulder, Colorado 80309-0596, United States
S Supporting Information
*
ABSTRACT: To reduce shrinkage stress which arises during
the polymerization of cross-linked polymers, allyl sulfide
functional groups were incorporated into methacrylate
polymerizations to determine their effect on stress relaxation
via addition−fragmentation chain transfer (AFCT). Addition-
ally, stoichiometrically balanced thiol and allyl sulfide-
containing norbornene monomers were incorporated into
the methacrylate resin to maximize the overall functional
group conversion and promote AFCT while also enhancing
the polymer’s mechanical properties. Shrinkage stress and
reaction kinetics for each of the various functional groups were measured by tensometry and Fourier-transform infrared (FTIR)
spectroscopy, respectively. The glass transition temperature (T ) and elastic moduli (E′) were measured using dynamic
g
mechanical analysis. When the allyl sulfide functional group was incorporated into dimethacrylates, the polymerization-induced
shrinkage stress was not relieved as compared with analogous propyl sulfide-containing resins. These analogous propyl sulfide-
containing monomers are incapable of undergoing AFCT while having similar chemical structure and cross-link density to the
allyl sulfide-containing methacrylates. Here, a monomethacrylate monomer that also contains a cyclic allyl sulfide (PAS) was
found to increase the cross-linking density nearly 20 times as compared to an analogous monomethacrylate in which the allyl
sulfide was replaced with an ethyl sulfide. Despite the much higher cross-link density, the PAS formulation exhibited no
concomitant increase in stress. Thiol−norbornene resins were copolymerized in PAS to promote AFCT as well as to
synergistically combine the ring-opening benefits associated with the thiol−ene reaction. AFCT resulted in a 63% reduction of
polymerization stress and a 45 °C enhancement of the glass transition temperature in the allyl sulfide-containing thiol−
norbornene−methacrylate system compared with rubbery dimethacrylates. When compared with conventional glassy
dimethacrylates, this combined system has less than 10% of the typical shrinkage stress level while having similarly excellent
mechanical properties.
6
7−9
INTRODUCTION
delayed, such as thiol−ene and thiol−yne
polymerizations,
■
have been used to enable the material to shrink in the early
stages of the reaction without leading to any significant stress
accumulation. Under these circumstances, prior to the gel
point, a material is capable of dissipating shrinkage via the
viscous flow of the monomer/oligomer species; however, after
gelation, molecular rearrangement is reduced and stress
accumulates. To address this issue, we have explored a
methodology utilizing reversible covalent bond structures to
Polymerization shrinkage stress is a deleterious phenomenon
1
−3
that plagues numerous polymer material applications
and
results in warping, microcracks, and material failure. Polymer-
ization shrinkage itself results from the decrease in molecular
spacing between monomers as the van der Waals interactions
are replaced with covalent bonds, bringing the monomer units
more closely together. Previously, various approaches have
been used to combat polymerization stress including
incorporation of materials that exhibit low polymerization
shrinkage, such as ring-opening monomers or formulations
that undergo polymerization-induced phase separation. Addi-
tionally, strategies that in which the gel-point conversion is
4
Received: March 18, 2012
Revised: June 15, 2012
Published: July 3, 2012
5
©
2012 American Chemical Society
5640
dx.doi.org/10.1021/ma300228z | Macromolecules 2012, 45, 5640−5646

Macromolecules
Article
enable the rearrangement of network strands. This approach
facilitates an elastic to plastic transition in behavior during the
polymerization that provides a route for postgel-point network
stress relaxation through accommodation of the shrinkage.
Specifically, we have utilized addition−fragmentation chain
transfer (AFCT) to enable nondegradative radical-mediated
breaking and re-forming of polymer network strands
10−12
throughout the polymerization.
The initial demonstrations
of reduced polymerization shrinkage stress using AFCT
employed the allyl sulfide functional group in both thiol−
1
1
13
ene and thiol−yne materials. These systems have the
additional advantages of possessing delayed gelation as well as
negligible oxygen inhibition.
6,14
While these systems demon-
strated decreased polymerization stress, the cross-linking
density of these step-growth polymers was necessarily lower
than their chain growth counterparts, leading to a lower cross-
link density, glass transition temperature, and modulus, which
precluded their use in many applications.
Methacrylate-based materials are widely used in a variety of
15
applications, such as electrical materials, optical materials, and
16
dental restorative materials, as they are typically lightweight,
possess a high modulus, exhibit good impact strength, and are
transparent. Unfortunately, the rapid polymerization of multi-
functional methacrylate (or acrylate) monomers generates a
Figure 1. Materials used: (1) SAS, (2) SPS, (3) PAS, (4) PES, (5)
NAS, (6) NPS, (7) PETMP, and (8) HCPK.
17
large amount of stress, on the order of megapascals, owing to
the typical chain-growth polymerization characteristics, partic-
ularly early gelation and the ultimate formation of a glassy,
component had been added. The flask contents were then refluxed for
4
5 min and cooled to ambient temperature. The white solid was
1
8,19
highly cross-linked polymer.
To achieve enhanced
removed by vacuum filtration, and the filter cake was washed with
excess ethanol. The solvent was removed in a rotary evaporator
followed by drying using a vacuum pump to give a colorless liquid. If
required, vacuum distillation (6−7 Torr) at 135−150 °C can be used
for purification of the crude product.
The synthesized 2-methylenepropane-1,3-bis(2-hydroxyethyl sul-
fide) (10.6 g, 0.051 mol) was dissolved in 100 mL of ethyl acetate in a
mechanical properties as well as the stress relaxation capabilities
of the AFCT mechanism, thiol−ene resins containing allyl
20
sulfide moieties were incorporated into dimethacrylate resins
to induce stress relaxation. As expected, the increased
methacrylate content enhances the mechanical properties
such as the glass transition temperature and elastic
5
00 mL three-neck round-bottom flask equipped with a dropping
2
1,22
modulus.
While stress relaxation was observed, the effect
funnel on one arm, a magnetic stirring bar, and a dry air blanket. To
the solution was added monomethacryloyloxyethylsuccinic acid (25 g,
0.108 mol, Sigma-Aldrich), followed by N,N-dimethylaminopyridine
(DMAP, 1.35 g, 0.011 mol, Alfa Aesar). The mixture was cooled in an
ice bath with continuous stirring. In a separate flask, dicyclohexyl
carbodiimide (DCC, Alfa Aesar) was dissolved in 150 mL of ethyl
acetate. The DCC solution was placed in the dropping funnel and
added dropwise to the flask as the flask contents were kept cold (0−10
was significantly reduced or even eliminated for materials
possessing a superambient glass transition temperature due to a
limited allyl sulfide concentration which was incorporated only
into the thiol−ene component of the resin.
In contrast, here, we incorporate the allyl sulfide functional
group directly into the methacrylate monomer to investigate its
effect on stress relaxation. In addition, tetrafunctional thiol and
allyl sulfide-containing difunctional norbornene monomers
were formulated into the methacrylate resin to enhance further
the mechanical properties and the capability for undergoing
AFCT.
°
C). After complete addition of the DCC solution, the flask contents
were continuously stirred in the ice bath for 30 min and then at
ambient temperature overnight. The following day, 50 mL of hexanes
was added, and all solids formed in the reaction flask were removed
and washed by vacuum filtration. The filtrate was then extracted with a
mixture of 100 mL of 1 N HCl, 150 mL of 10% aqueous sodium
bicarbonate, and 100 mL of water. The organic layer was dried
(Na SO ) and then concentrated in a rotary evaporator followed by
EXPERIMENTAL SECTION
■
2
4
Materials. Chemical structures of materials utilized in this study are
shown in Figure 1.
drying to give a light yellow liquid in 90% yield. The structure was
confirmed by 1H NMR spectroscopy (Varian INOVA 500) (see
Supporting Information).
Synthesis of 2-Methylenepropane-1,3-bis(2-[2-methacryloy-
loxyethyl succinyl]ethyl sulfide) (SAS, Succinate Allyl Sulfide).
To synthesize the desired molecular precursor, 2-methylenepropane-
Synthesis of 2-Methylpropane-1,3-bis(2-[2-methacryloylox-
yethyl succinyl]ethyl sulfide) (SPS, Succinate Propyl Sulfide).
To synthesize 2-methylpropane-1,3-bis(2-hydroxyethyl sulfide), crude
2-methylpropane-1,3-bis(2-hydroxyethyl sulfide) was prepared from 2-
methyl-1-bromo-3-chloropropane (9.57 g, 0.169 mol, Sigma-Aldrich),
mercaptoethanol (26.44 g, 0.339 mol, TCI America), and sodium
metal (8.6 g, 0.374 mol, Lancaster Synthesis Inc.) in a method similar
to the one described for methylenepropane-1,3-bis(2-hydroxyethyl
sulfide) above. Vacuum distillation at 6−8 Torr and 140−160 °C gave
a colorless liquid with 77.5% recovery.
1,3-bis(2-hydroxyethyl sulfide), 2-mercaptoethanol (25.60 g, 0.328
mol, Alfa Aesar) was dissolved in 100 mL of ethanol in a 500 mL two-
neck round-bottom flask equipped with a magnetic stirring bar,
condenser, and dropping funnel. With vigorous stirring, sodium metal
(
8.20 g, 0.356 mol, Alfa Aesar) was added slowly in small pieces to
minimize the reaction exotherm. After complete addition of the
sodium metal, the mixture was stirred under a nitrogen blanket until
the flask temperature returned to ambient. A solution of 3-chloro-2-
chloromethyl-1-propene (20 g, 0.16 mol, Secant Chemicals Inc.) in 50
mL of ethanol was added dropwise to give a white cloudy mixture that
transitioned to a white solid by the time all the dichloropropene
The mono-HEMA succinic acid intermediate was made in situ and
then used in the second step in a one-pot fashion. 2-Hydroxyethyl 2-
methylprop-2-enoate (HEMA, 10 g, 0.084 mol, Sigma-Aldrich) and
5
641
dx.doi.org/10.1021/ma300228z | Macromolecules 2012, 45, 5640−5646

Macromolecules
Article
succinic anhydride (9.35 g, 0.094 mol, Sigma-Aldrich) were charged
into a round-bottom flask equipped with a magnetic stirring bar, a dry
air blanket, and dropping funnel. DMAP (1.15 g, 0.009 mol) and
butylated hydroxytoluene (BHT, 0.035 g, 0.16 mmol) were added to
the flask. The mixture was heated with stirring at 90 °C for 5 h to give
a colorless, viscous liquid, and the flask was left to cool to ambient
temperature. Ethyl acetate (∼50 mL) was subsequently added, and the
mixture was cooled in an ice bath to 0−5 °C. Charged into the
dropping funnel was a solution of DCC (17.3 g, 0.089 mol, Alfa Aesar)
in 50 mL of ethyl acetate. The DCC solution was carefully and slowly
added to the vigorously stirred cold solution over 30 min, and the
mixture was stirred at 0−5 °C for 30 min then at ambient temperature
overnight. The following day, the reaction was worked up similarly to
cantilever beam deflection that is detected by the LVDT (linear
variable displacement transducer) displacement. In the tensometer,
one glass rod is fixed to the bottom plate and another rod is connected
to the cantilever beam. The formulated resin was injected between the
two glass rods, which results in a specimen geometry of 6 mm
diameter and 1 mm thickness. Samples were irradiated with 365 nm
2
filtered UV light at 10 or 50 mW/cm (Acticure 4000, EXPO) for 5 or
16 min. PAS and PES samples were exposed to the higher light
intensity because they polymerize much more slowly otherwise than
the remainder of the monomers considered here. Evolution of the
methacrylate, norbornene and allyl sulfide double bond concentrations
were determined by monitoring the infrared absorption peaks centered
−
1
−1
at 6164 cm (CC−H stretching, overtone), 6111 cm (CC−H
1
−1
SAS above, and the structure was confirmed by H NMR spectroscopy
stretching, overtone), and 6121 cm (CC−H stretching, over-
(
Varian INOVA 500) (see Supporting Information).
tone), respectively. As the methacrylate peak is overlapped with the
norbornene and allyl sulfide peaks, Gaussian fitting was used to
deconvolute these peak areas. See the Supporting Information for
additional detail on the infrared spectra and the conversion
measurement.
Synthesis of 2-(Methacryloyloxyethyl)-7-methylene-1,5-di-
thiocan-3-yl Phthalate (PAS, Phthalate Allyl Sulfide) and 2-
(
(
Methacryloyloxyethyl)-7-methyl-1,5-dithiocan-3-yl Phthalate
PES, Phthalate Ethyl Sulfide). PAS and PES were synthesized
2
3
following the procedure described in the literature. Mono-2-
methacryloyloxyethyl phthalate (9.4 g, 31 mmol, Sigma-Aldrich) and
C-8 alcohol (5.38 g, 31 mmol), DMAP (400 mg), DCC (6.95 g, 34
mmol, Alfa Aesar), and methylene chloride (50 mL) were utilized. The
structures of products were confirmed by 1H NMR spectroscopy
The elastic moduli (E′) and glass transition temperatures (T s) of
g
polymerized samples were measured by dynamic mechanical analysis
(DMA, TA Instruments Q800). Specimens (25 mm × 5 mm × 1 mm)
for DMA were prepared and irradiated under the identical conditions
as performed in the tensometer experiment. DMA experiments were
performed at a constant strain and frequency of 0.1% and 1 Hz,
respectively, scanning the temperature from −50 to 140 °C twice at 1
(
Varian INOVA 500) (see Supporting Information).
Synthesis of 2-Methylenepropane-1,3-di(norbornene sul-
fide) (NAS, Norbornene Allyl Sulfide). 5-Bromomethylnorbornene
was synthesized from 1,3-dicyclopentadiene (16 g, 0.12 mol, ACROS),
allyl bromide (35 g, 0.29 mol, Sigma-Aldrich), and hydroquinone (81
mg, 0.74 mmol, Aldrich) by adding them in a 100 mL pressure vessel
followed by heating at 170 °C for 12 h. The crude oil was purified with
normal distillation at 175 °C. 3-Mercapto-2-(mercaptomethyl)-1-
propene was synthesized according to the method described in the
°
C/min; the temperature scan was repeated to ensure the absence of
dark polymerization at temperatures greater than the Tg.
27,28
RESULTS AND DISCUSSION
■
The allyl sulfide functional group was incorporated into two
different monomers containing methacrylate functionalities: a
symmetric dimethacrylate with a central allyl sufide functional
group (succinate-based allyl sulfide, SAS) and an asymmetric
monomethacrylate with an allyl sulfide containing ring that
participates in the polymerization through a ring-opening
reaction (phthalate-based allyl sulfide, PAS). Both monomers
were designed to undergo radical-mediated AFCT during
polymerization while PAS has the ring-opening functional
group for further stress reduction and the stiff benzene ring
structure to promote a higher glass transition temperature than
SAS. To isolate the effects of AFCT on stress relaxation,
analogous monomers were synthesized that possess a nearly
identical molecular structure; however, each “control” mono-
mer does not contain an allyl sulfide and is thus incapable of
undergoing AFCT. These monomers were created such that, to
the greatest extent possible, differences in the polymerization
stress development could be attributed to the plasticity enabled
by the AFCT mechanism for stress relaxation. Specifically, the
allyl sulfides in SAS and PAS were replaced by propyl sulfide
1
1,24,25
literature.
of 5-bromomethylnorbornene and 10.3 g (86.2 mmol) of 2-methyl-
,3-dimercaptopropene in methanol was prepared and added to a
Then, a slightly diluted solution of 19 g (172.4 mmol)
1
refluxing solution of sodium (5 g, 215.5 mmol) in 300 mL of methanol
under argon protection. The following day, methanol was dried under
vacuum, and the dried crude oil was purified using liquid−liquid
water−ether solution. The product extracted in the ether phase was
dried and vacuum-distilled at 230 °C and 0.3 mmHg to give a waxy/
1
hazy solid. The structure was confirmed by H NMR spectroscopy
(
Varian INOVA 500) (see Supporting Information).
Synthesis of 2-Methylpropane-1,3-di(norbornene sulfide)
NPS, Norbornene Propyl Sulfide). NPS was synthesized from 5-
(
bromomethylnorbornene with 1,3-dimercapto-2-methylpropane with
the same synthetic procedure of NAS. 3-Dimercapto-2-methylpropane
11
was synthesized following the procedure in the literature. A viscous/
colorless liquid (NPS) was given after vacuum purification at 200 °C
1
and 0.3 mmHg, and its structure was confirmed by H NMR
spectroscopy (Varian INOVA 500) (see Supporting Information).
Tetrathiol (pentaerythritol tetrakis(3-mercaptopropionate)
(
(
PETMP, Evans Chemetics), 1-hydroxycyclohexylphenyl ketone
HCPK, Ciba Specialty Chemicals), and sodium metal (Sigma-
(
SPS) or ethyl sulfide (PES), respectively.
The comparison between the conversion and stress evolution
Aldrich) were provided or purchased, and no further purification was
performed. All resin mixtures were produced based on a desired
functional group ratio. For SAS−PETMP, SPS−PETMP resins, the
functional group ratio (methacylate:thiol) was 7 to 3. NAS−PETMP−
PAS and NPS−PETMP−PAS were stoichiometically balanced,
implying that the ratio of norbornene to thiol to methacrylate was
in the methacrylate monomers is shown in Figure 2. When the
allyl sulfide was incorporated into the dimethacrylate-based
monomer (SAS), the stress evolution is nearly identical when
compared with SPS, indicating that the AFCT mechanism is
not promoting plasticization under these circumstances. A
possible explanation for this behavior is that the allyl sulfide
exhibits reduced reactivity toward carbon-centered radicals such
as those involved in the methacrylate polymerization. Further,
the reaction of a carbon-centered radical with the allyl sulfide is
not a reversible reaction as the product contains an allylic group
that is distinct from the initial allyl sulfide (Scheme 1B).
Ultimately, both of these explanations would lead to reduced
capability for undergoing chain cleavage and re-formation, as is
1:1:1. Throughout this study, all resins includes 1 wt % of HCPK as a
UV-activated photoinitiator.
Methods and Equipment. The shrinkage stress and functional
group conversion were simultaneously observed during polymerization
using a tensometer coupled with a FTIR spectrometer (Nicolet 670),
which was equipped with near-infrared transmitting optical fiber patch
6
,26
cables and an indium gallium arsenide (InGaAs) detector.
tensometer was developed by the Paffenbarger Research Center
American Dental Association Health Foundation) to measure the
stress during photopolymerization. Stress evolution is measured by the
The
(
5
642
dx.doi.org/10.1021/ma300228z | Macromolecules 2012, 45, 5640−5646

Macromolecules
Article
containing resins compared with the propyl sulfide-containing
resins (Figure 2B) although stresses of both systems were
reduced compared with those of SAS or SPS due to the
addition of PETMP. This outcome is attributed to the
enhanced cross-linking density of the SAS-PETMP resin that
results from the higher conversion of the allyl sulfide functional
group (33%), which, under these circumstances, leads to
additional cross-linking in the allyl sulfide-containing system.
The shrinkage stress of the phthalate allyl sulfide-based
methacrylate (PAS) resin is lower than in the succinate allyl
sulfide-based methacrylate resin (SAS). However, as shown in
Figure 2B, the phthalate ethyl sulfide methacrylate (PES)
demonstrated similar stress evolution behavior to the allyl
sulfide-containing material (PAS) which serves as a molecular
control. Despite serving as a molecular control, the PES by its
nature is significantly different both in how it polymerizes and
in enabling AFCT. While the molecular structures of the two
monomers (PAS and PES) are nearly identical, the allyl sulfide
functional group in this monomer uniquely acts both to
promote AFCT and as an additional polymerizable functional
group through a ring-opening reaction. This ring-opening
reaction occurs during the first allyl sulfide AFCT reaction of
the monomer and leads to cross-linking in this material. Thus,
the PES monomer is not capable of AFCT but also not capable
of cross-linking, leading to dramatic differences in the cross-link
density, modulus, and inherent shrinkage stress that will arise.
These effects are clearly seen in the material properties for
these two resins (see Figure 3) as well, where PES does not
exhibit a clearly defined rubbery modulus, but instead the
behavior more closely resembles that of a very loosely cross-
linked polymer melt. This outcome suggests that, as expected,
the ethyl sulfide ring in the PES structure does not undergo
AFCT but also does not significantly react to form cross-links.
Nevertheless, it should be noted that the two materials have
nearly identical polymerization stresses, indicating that the
additional cross-linking and higher modulus resulting from the
cross-linking of the allyl sulfide monomer are counterbalanced
by AFCT. Overall, this combination leads to enhanced
mechanical properties and behavior with no concomitant stress
increase.
Figure 2. (A) Methacrylate conversion and (B) shrinkage stress versus
time for PAS (□) (allyl sulfide-based monomethacrylate), PES (■)
(
ethyl sulfide-based monomethacrylate), SAS (○) (allyl sulfide-based
dimethacrylate), SPS (●) (propyl sulfide-based dimethacrylate), SAS-
PETMP (△), and SPS-PETMP (▲). Samples were formulated with 1
wt % HCPK as the photoinitiator. All resins except PAS and PES (50
2
2
mW/cm ) were irradiated with 365 nm light at 1 mW/cm for 16 min.
Scheme 1. (A) Reversible AFCT Is Mediated by Reaction
with the Thiyl Radical While (B) Irreversible Allyl Sulfide
AFCT Is Mediated by Reaction with a Carbon-Centered
Radical
As discussed above, the addition of thiol−ene monomers
into a methacrylate resin can have the effect of enhancing stress
reduction by promoting allyl sulfide-mediated AFCT in the
20
methacrylate resins. As our goal is to have a low-stress
material possessing a superambient T , a stoichiometrically
g
balanced mixture of thiol−ene monomers (ene functional
group:thiol functional group = 1:1) was incorporated into the
phthalate-based methacrylate (PAS), since its T is significantly
g
higher than those of the SAS and SPS resins. Further, to
maximize the amount of allyl sulfide and thus the AFCT
reaction in the cross-linked network, a dinorbornene (i.e.,
“
diene”) monomer was synthesized. The dinorbornene allyl
sulfide (NAS) monomer was compared with a propyl sulfide-
containing norbornene (NPS) to isolate the stress relaxation
effect of the allyl sulfide functional group. Thus, by utilizing
PAS with thiol-NAS and thiol-NPS, the effect of the allyl
sulfide’s concentration in this ternary resin is elucidated. In each
of the two combinations presented here, the norbornene, thiol,
and methacrylate functional group were stoichiometrically
balanced (norbornene:thiol:methacrylate = 1:1:1), though it
is not necessary to maintain a 1:1 ratio of methacrylate to
thiol−ene formulation.
necessary to promote stress relaxation. These explanations
would also result in a low conversion of the allyl sulfide in SAS
(
Table 1) and lead to SAS and SPS exhibiting similar
mechanical properties (Figure 3), as observed. To explore
this possibility further, a tetrafunctional thiol (PETMP) was
formulated into these resins to introduce thiyl radicals, which
have been shown to have excellent reactivity toward allyl
sulfides and thus promote AFCT in a fully reversible manner, as
11
shown in Scheme 1A. Interestingly, the addition of PETMP
resulted in an enhanced shrinkage stress for the allyl sulfide-
5
643
dx.doi.org/10.1021/ma300228z | Macromolecules 2012, 45, 5640−5646

Macromolecules
Article
a
Table 1. Summary of the Final Functional Group Conversions, T , and Elastic Moduli
g
final functional group conversion (%)
vinyl ether monomer used in the ternary
system (thiol:vinyl ether:methacrylate
ratio)
rubbery elastic modulus
allyl
sulfide
at T = T + 40 °C
elastic modulus at
T = 25 °C (MPa)
g
methacrylate norbornene
stress (MPa)
Tg (°C)
(MPa)
SAS
SPS
PAS
PES
84 ± 0.8
92 ± 0.5
77 ± 0.2
82 ± 0.7
10 ± 1.5
16 ± 0.2
0.55 ± 0.004 24 ± 1
0.57 ± 0.004 20 ± 1
19 ± 0.2
18 ± 0.1
35 ± 0.3
2 ± 0.8
86 ± 7
55 ± 2
2490 ± 500
1870 ± 300
0.36 ± 0.01
0.33 ± 0.01
102 ± 0.7
below
8
0 ± 4
NAS−PETMP−PAS
NPS−PETMP−PAS
100 ± 0.1
100 ± 0.1
64 ± 0.5
70 ± 0.8
21 ± 0.9
5 ± 0.2
0.21 ± 0.01
0.28 ± 0.01
65 ± 1
65 ± 2
19 ± 0.5
14 ± 0.4
2140 ± 400
1900 ± 400
a
Rubbery moduli are measured at the same temperature for both the allyl sulfide-based resin and its analogous non-allyl sulfide-containing resin.
When the T s are different for these formulations, the rubbery modulus is taken at the higher of the two glass transition temperatures plus 40 °C.
g
Figure 3. (A) Elastic modulus and (B) tan δ versus temperature for
PAS (□) (allyl sulfide-based monomethacrylate), PES (■) (ethyl
sulfide-based monomethacrylate), SAS (○) (allyl sulfide-based
Figure 4. (A) Methacrylate conversion and (B) shrinkage stress versus
dimethacrylate), and SPS (●) (propyl sulfide-based dimethacrylate).
time for stoichiometric mixtures of NAS−PETMP−PAS (○)
Samples were formulated with 1 wt % HCPK. SAS and SPS were
(
functional group ratio norbornene:thiol:methacrylate = 1:1:1) and
2
irradiated with 365 nm light at 1 mW/cm for 16 min. PAS and PES
NPS−PETMP−PAS (●) (functional group ratio norbornene:thiol:-
2
were irradiated with 365 nm light at 50 mW/cm for 16 min.
methacrylate = 1:1:1). Samples were formulated with 1 wt % HCPK
2
and irradiated with 365 nm light at 50 mW/cm for 5 min.
The reaction kinetics in these resins exhibited similar
behavior whether the allyl sulfide- or propyl sulfide-based
norbornene was included, where the methacrylate functional
group proceeds to nearly complete conversion at a more rapid
rate than the thiol−ene component (Figure 4A). A relatively
low final norbornene conversion was observed, likely resulting
from the reduced molecular mobility associated with
vitrification (Table 1) that arises at the end of these
polymerizations. The allyl sulfide conversion was limited to at
most 20% conversion (Table 1), indicative of a largely
reversible AFCT reaction as desired to minimize the stress
that develops. In particular, the allyl sulfide functional group
was only negligibly consumed when present only in the
methacrylate monomer (i.e., the NPS−PETMP−PAS resin). As
the rubbery modulus of the PAS is much greater than PES (i.e.,
without the ring-opening component), the allyl sulfide is being
incorporated into the network via ring-opening, but it is not
being consumed by irreversible side reactions. Since the AFCT
mechanism results in ring-opening while also preserving the
allyl sulfide functional group (see Scheme 1A), we conclude
that stress relieving AFCT must occur subsequent to the ring-
opening reaction in order to promote plasticity and stress
5
644
dx.doi.org/10.1021/ma300228z | Macromolecules 2012, 45, 5640−5646

Macromolecules
Article
relaxation during the polymerization. This outcome is further
evidenced by NAS−PETMP−PAS having lower polymerization
stress than NPS−PETMP−PAS, which has a reduced allyl
sulfide content in the formulation (i.e., NPS versus NAS). Since
these two ternary resins form very similar polymer networks as
demonstrated in Figure 5, the reduced stress in the NAS−
than their high stress counterparts. Comparisons with conven-
tional high T dimethacrylates are even more astonishing as the
g
NAS−PETMP−PAS system demonstrated more than a factor
of 10 lower stress level as compared with that previously
6
reported for conventional dimethacrylates (2.9 MPa) while
simultaneously achieving excellent mechanical properties. Thus,
this resin system is an outstanding candidate for applications in
which the stress level is critical such as electronic materials,
optical materials, and dental materials.
CONCLUSIONS
■
Stress relaxation during the photopolymerization of monomers
containing allyl sulfide functional groups was evaluated in
methacrylate and thiol−norbornene−methacrylate systems.
The allyl sulfide’s capability to undergo AFCT resulted in a
decrease in polymerization stress when compared with
analogous materials incapable of undergoing AFCT while also
having similar cross-link densities. Although AFCT did not
relieve stress when the allyl sulfide functionality was
incorporated directly into dimethacylates, the ring-opening
allyl sulfide-based methacrylate (PAS) increased the cross-
linking density nearly 20 times and resulted in a T increase of
g
up to 20 °C without a concomitant increase in stress as
compared with the ethyl sulfide-based methacrylate (PES). PAS
was then copolymerized in a thiol−norbornene resin (NAS−
PETMP) to facilitate AFCT as well as to incorporate the
benefits associated with the thiol−ene reaction. The allyl sulfide
functional group in the NAS−PETMP−PAS system resulted in
a factor of 3 lower stress as compared to the dimethacrylate
resin (SPS) while achieving a higher T by up to 40 °C and
g
excellent conversion. PAS and NAS−PETMP−PAS systems
demonstrate more than a factor of 10 lower stress as compared
6
with conventional dimethacrylates while also exhibiting
excellent mechanical properties.
ASSOCIATED CONTENT
Supporting Information
Figures S1−S4. This material is available free of charge via the
Internet at http://pubs.acs.org.
■
*
S
AUTHOR INFORMATION
Corresponding Author
E-mail: christopher.bowman@colorado.edu.
■
Figure 5. (A) Elastic modulus and (B) tan δ versus temperature for
stoichiometrically balanced mixtures of NAS−PETMP−PAS (○)
*
Notes
(
functional group ratio norbornene:thiol:methacrylate = 1:1:1) and
NPS−PETMP−PAS (●) (functional group ratio norbornene:thiol:-
The authors declare no competing financial interest.
methacrylate = 1:1:1). Samples were formulated with 1 wt % HCPK
2
and irradiated with 365 nm at 50 mW/cm for 5 min.
ACKNOWLEDGMENTS
This investigation was supported by NIDCR 2 R01 DE-
10959-11 from the National Institutes of Health and NSF
933828.
■
0
0
PETMP−PAS system results from stress relaxation by AFCT.
In addition, NAS−PETMP−PAS has a higher cross-link density
(
Table 1), which in the absence of AFCT would lead to larger
REFERENCES
polymerization stress. The higher stress is not observed because
of the AFCT-induced rearrangement of the network.
■
(1) Drury, C. J.; Mutsaers, C. M. J.; Hart, C. M.; Matters, M.; de
Leeuw, D. M. Appl. Phys. Lett. 1998, 73, 108−110.
2) Bowman, C. N.; Cramer, N. B.; Stansbury, J. W. J. Dent. Res.
Amazingly, despite the formation of glassy, cross-linked
polymers that would be expected to limit the amount of
relaxation, this approach to stress reduction was highly
successful at alleviating stress relative to almost any other
approach. In particular, the allyl sulfide-incorporated norbor-
nene−PETMP−PAS systems demonstrated 63% reduced stress
as compared to the dimethacrylate resin (SPS) while achieving
higher functional group conversions and exhibiting glass
transition temperatures that were as much as 40 °C higher
(
2
(
(
011, 90, 402−416.
3) Yoffe, A. D. Adv. Phys. 2001, 50, 1−208.
4) Alcoutlabi, M.; McKenna, G. B.; Simon, S. L. J. Appl. Polym. Sci.
2
(
003, 88, 227−244.
5) Lu, H.; Trujillo-Lemon, M.; Ge, J.; Stansbury, J. W. Compend.
Contin. Educ. Dent. 2010, 31 Spec No. 2, 1−4.
(6) Lu, H.; Carioscia, J. A.; Stansbury, J. W.; Bowman, C. N. Dent.
Mater. 2005, 21, 1129−1136.
5
645
dx.doi.org/10.1021/ma300228z | Macromolecules 2012, 45, 5640−5646

Macromolecules
Article
(
7) Chan, J. W.; Shin, J.; Hoyle, C. E.; Bowman, C. N.; Lowe, A. B.
Macromolecules 2010, 43, 4937−4942.
8) Fairbanks, B. D.; Scott, T. F.; Kloxin, C. J.; Anseth, K. S.;
Bowman, C. N. Macromolecules 2009, 42, 211−217.
9) Lowe, A. B.; Hoyle, C. E.; Bowman, C. N. J. Mater. Chem. 2010,
0, 4745−4750.
10) Kloxin, C. J.; Scott, T. F.; Adzima, B. J.; Bowman, C. N.
Macromolecules 2010, 43, 2643−2653.
11) Kloxin, C. J.; Scott, T. F.; Bowman, C. N. Macromolecules 2009,
2, 2551−2556.
12) Scott, T. F.; Schneider, A. D.; Cook, W. D.; Bowman, C. N.
Science 2005, 308, 1615−1617.
13) Park, H. Y.; Kloxin, C. J.; Scott, T. F.; Bowman, C. N.
Macromelecules 2010, 43, 10188−10190.
14) Carioscia, J. A.; Lu, H.; Stanbury, J. W.; Bowman, C. N. Dent.
(
(
2
(
(
4
(
(
(
Mater. 2005, 21, 1137−1143.
(15) Kloosterboer, J. G. Adv. Polym. Sci. 1988, 84, 1−61.
(16) Peutzfeldt, A. Eur. J. Oral Sci. 1997, 105, 97−116.
(17) Kleverlaan, C. J.; Feilzer, A. J. Dent. Mater. 2005, 21, 1150−
1
157.
18) Labella, R.; Lambrechts, P.; Van Meerbeek, B.; Vanherle, G.
Dent. Mater. 1999, 15, 128−137.
19) Walls, A. W. G.; Mccabe, J. F.; Murray, J. J. J. Dent. (Oxford, U.
K.) 1988, 16, 177−181.
20) Park, H. Y.; Kloxin, C. J.; Scott, T. F.; Bowman, C. N. Dent.
Mater. 2010, 26, 1010−1016.
21) Lee, T. Y.; Carioscia, J.; Smith, Z.; Bowman, C. N.
Macromolecules 2007, 40, 1473−1479.
22) Lee, T. Y.; Smith, Z.; Reddy, S. K.; Cramer, N. B.; Bowman, C.
N. Macromolecules 2007, 40, 1466−1472.
23) Abuelyaman, A. S.; M, S. B.; Lewandowski, K. M.; Plaut, D. J.
(
(
(
(
(
(
Dental Compositions Containing Hybrid Monomers; 3M Innovative
Properties Company: St. Paul, MN, 2011.
(
24) Evans, R. A.; Rizzardo, E. Macromolecules 2000, 33, 6722−6731.
(
25) Evans, R. A.; Rizzardo, E. J. Polym. Sci., Polym. Chem. 2001, 39,
2
(
02−215.
26) Lu, H.; Stansbury, J. W.; Dickens, S. H.; Eichmiller, F. C.;
Bowman, C. N. J. Mater. Sci.: Mater. Med. 2004, 15, 1097−1103.
27) Ferrillo, R. G.; Achorn, P. J. J. Appl. Polym. Sci. 1997, 64, 191−
95.
28) Li, G.; Lee-Sullivan, P.; Thring, R. W. J. Therm. Anal. Calorim.
000, 60, 377−390.
(
1
(
2
5
646
dx.doi.org/10.1021/ma300228z | Macromolecules 2012, 45, 5640−5646