Effect of Segmental Configuration on Properties of n-Propyl-Bridged Polycyanurate Networks

Article abstract of DOI:10.1021/acs.macromol.7b00824

The effect of the chemical configuration of network segments on the physical properties, cure properties, mechanical performance, and chemical stability of polycyanurate networks was investigated via synthesis, network formation, and characterization of an isomeric series of n-propyl-bridged cyanate ester monomers. Configurations that provide cyanurate oxygen atoms with either nearby methyl groups or nearby bridge groups exhibited decreased moisture uptake by up to 50%, along with a roughly 20-40 °C reduction in the loss in glass transition temperatures due to hydrolysis, for networks immersed in 85 °C water for 96 h. In ortho,para-linked aryl cyanates, dry glass transition temperatures of cured networks were reduced compared to analogous para,para-linked networks by only about 10 °C, compared to a reduction of 30 °C in ortho-methylated cyanate ester networks, leading to higher wet glass transition temperatures in the ortho,para-linked networks. Neither methyl groups nor bridge groups in a position ortho to the reactive cyanate ester groups prevented the creation of networks with >99% conversion at cure temperatures of 230 °C. Networks with placement of methyl groups in a position ortho to the cyanate ester exhibited char yields in nitrogen at 600 °C of 46-47% compared to 43% for networks with methyl groups in the corresponding meta position, regardless of whether a sterically hindered environment was present around the cyanurate oxygen. These results illustrate the manner in which the chemical configuration around reactive groups can substantially modify the properties of networks even when the number density and type of reactive group present do not change.

Full text of DOI:10.1021/acs.macromol.7b00824

Article
pubs.acs.org/Macromolecules
Effect of Segmental Configuration on Properties of n‑Propyl-Bridged
Polycyanurate Networks
,
†
§
§
§
‡
Andrew J. Guenthner,* Benjamin G. Harvey, Andrew P. Chafin, Matthew C. Davis, Jacob J. Zavala,
‡
‡
†
†
Kevin R. Lamison, Josiah T. Reams, Kamran B. Ghiassi, and Joseph M. Mabry
†
‡
Aerospace Systems Directorate and ERC Incorporated, Air Force Research Laboratory, Edwards AFB, California 93524, United
States
§
Weapons Division, Naval Air Warfare Center, China Lake, California 93555, United States
S Supporting Information
*
ABSTRACT: The effect of the chemical configuration of
network segments on the physical properties, cure properties,
mechanical performance, and chemical stability of polycyanu-
rate networks was investigated via synthesis, network
formation, and characterization of an isomeric series of n-
propyl-bridged cyanate ester monomers. Configurations that
provide cyanurate oxygen atoms with either nearby methyl
groups or nearby bridge groups exhibited decreased moisture
uptake by up to 50%, along with a roughly 20−40 °C reduction
in the loss in glass transition temperatures due to hydrolysis,
for networks immersed in 85 °C water for 96 h. In ortho,para-linked aryl cyanates, dry glass transition temperatures of cured
networks were reduced compared to analogous para,para-linked networks by only about 10 °C, compared to a reduction of 30
°
C in ortho-methylated cyanate ester networks, leading to higher “wet” glass transition temperatures in the ortho,para-linked
networks. Neither methyl groups nor bridge groups in a position ortho to the reactive cyanate ester groups prevented the creation
of networks with >99% conversion at cure temperatures of 230 °C. Networks with placement of methyl groups in a position
ortho to the cyanate ester exhibited char yields in nitrogen at 600 °C of 46−47% compared to 43% for networks with methyl
groups in the corresponding meta position, regardless of whether a sterically hindered environment was present around the
cyanurate oxygen. These results illustrate the manner in which the chemical configuration around reactive groups can
substantially modify the properties of networks even when the number density and type of reactive group present do not change.
INTRODUCTION
An important prerequisite for developing structure−property
relationships is the availability of a variety of carefully controlled
and varied chemical structures, with quantifiable structural
characteristics. Careful control of the chemical structure enables
the avoidance of problems such as a network with too high of a
■
As a class of thermosetting polymer networks, polycyanu-
1
−4
rates
have proven to be exceptionally useful in extreme
environments due to the highly stable nature of the cyanurate
cross-links formed from cyanate ester monomers. Some notable
environments where polycyanurate networks have proven
useful include the charged particle and radiation environment
around Jupiter encountered by the Juno space probe, the heat
of atmospheric entry for the Mars Curiosity heat shield, and
glass transition temperature (T ) at full cure. Such networks
G
typically cannot be cured completely without chemical
degradation due to the high cure temperatures required, thus
making it impossible to measure the properties of the fully
5
6
12
cross-linked structure. At the same time, large variations in
structure must also be achievable; otherwise, the structure−
property correlations are likely to become unduly biased by a
specific subset of chemical features appearing in all the
structures.
the intense neutron bombardment expected within the ITER
thermonuclear fusion reactor
7
−9
and other superconducting
10
magnets exposed to high radiation doses. From a more
fundamental perspective, the high stability of the cyanurate
cross-link allows for the formation of networks with a very well-
1
3,14
Recently, bio-based materials such as anethole,
resvera-
1
1
15,16
17,18
19
20
21
defined and, fortuitously, easy to monitor chemical structure
even in the fully cross-linked state. This feature, along with the
simplicity of network formation by straightforward heating of
the monomer with very little net shrinkage or volatile
generation, makes polycyanurate networks ideal candidates
for studying the general physics of highly cross-linked polymer
networks and establishing useful structure−property relation-
ships.
trol,
eugenol,
cresol, vanillin, and turpentine have
been utilized to synthesize cyanate ester and other high-
22−30
temperature network
monomers with a much wider
variety of chemical structures than found in the typical
Received: April 20, 2017
Revised: June 11, 2017
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
petroleum-derived bisphenol or novolac monomers. The wider
variety of structures has facilitated systematic structural
comparisons, such as the effect of a methoxy group ortho to
the cyanate ester group, on the properties of networks. With
cresol-based monomers in particular, the bridge type, cyanate
ester configuration, and methyl group configuration in a bis-
phenyl motif may all be varied systematically. Figure 1
illustrates this concept using four isomeric chemical structures.
substantial bridge group. As shown in Figure 2, in this
configuration, the bridge itself may provide some additional
steric hindrance around the cyanate ester group ortho to the
bridge linkage. The ortho,para-linked monomers are typical side
products of para,para-linked monomer synthesis. In particular,
when m-cresol is used as a starting material, the formation of
the ortho,para-linked monomer may be enhanced by steric
effects. If these ortho,para-linked structures were to impart a
useful property such as lower moisture uptake, then it may be
possible to produce superior networks from a broader variety of
starting materials, with less demanding synthesis and
purification requirements.
31
In this paper, we report on a systematic study of n-propyl-
bridged, bis-phenyl cyanate ester monomers that can be derived
from bio-based starting ingredients. The n-propyl bridge is large
enough to contribute to steric hindrance, while simultaneously
lowering the glass transition temperature of the network at full
cure sufficiently to avoid issues with degradation, but not so
large as to unduly compromise thermal and mechanical
performance. We show that the use of the ortho,para-
substituted monomer is effective at lowering moisture uptake
while preserving other key properties. Through this carefully
selected systematic study, we therefore validate a key general
principle for minimizing moisture uptake in densely cross-
linked polymer networks.
Figure 1. Illustration of a set of four isomeric bis(cyanate ester)
monomeric compounds derived from m-cresol (1 and 3) or o-cresol (2
and 4) in which the methyl group configuration (compare 1 and 2),
cyanate ester configuration (compare 1 and 3), and bridge group
EXPERIMENTAL SECTION
■
(
compare 2 and 4) are all varied. Note that some properties for 1
31
Materials. Reagents and solvents were obtained commercially and
used as received unless otherwise specified.
prepared from creosol and its networks as well as 4 and its
networks have been previously published.
32
1
13
General Considerations. H and C NMR data were obtained
1
13
on a Bruker Avance II 300 MHz spectrometer ( H at 300 MHz, C at
5 MHz) using CDCl3, acetone-d , or DMSO-d and spectra
Previous studies of polycyanurate networks have revealed
that methyl group configuration can have a significant impact
on key physical properties of polycyanurate networks, such as
7
6
6
referenced against residual solvent or tetramethylsilane. Spectra are
available in the Supporting Information. Mass spectral data were
obtained on an Agilent 6890/5973N gas chromatograph system
(details in the Supporting Information). Elemental analysis was
conducted by Atlantic Microlab, Inc. (Norcross, GA).
Monomer Synthesis. 4,4′-(Propane-1,1-diyl)bis(3-methylphe-
nol) (5) and 2-(1-(4-Hydroxy-2-methylphenyl)propyl)-5-methyl-
pheol (7). A mechanically stirred mixture of 144 mL of
propionaldehyde (2 mol) and 840 mL of m-cresol (8 mol) was
cooled to −10 °C while 4 mL of concentrated HCl was added over 30
min. After an additional 30 min the cooling bath was removed, and the
mixture was stirred for 2 h. Excess m-cresol was then removed by
heating the mixture under high vacuum (0.3 Torr) until the internal
3
2,33
moisture uptake.
In particular, steric hindrance around the
cyanate ester group is believed to reduce moisture uptake,
although it may also interfere with cure chemistry. Because the
cure reactions for cyanate ester monomers may be initiated
with a proton source, and because of the ability for protons to
fit easily into spaces where water molecules do not fit, it should
be possible to identify forms of steric hindrance that prevent
water uptake without unduly sacrificing ease of cure. One very
interesting such candidate is the use of an ortho,para-linked
cyanate ester configuration in a bis-phenyl monomer with a
Figure 2. Illustration of various forms of steric hindrance found in the repeat structures of polycyanurate networks: (a) a bulky bridge group
highlighted yellow) may hinder access to a cyanurate oxygen atom in a manner similar to (b) a methyl group (highlighted orange) ortho to a
cyanurate linkage. Color scheme, C: gray; H: white; N: blue; O: red. For simplicity, repeat unit structures are H-terminated.
(
B
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
temperature rose above 90 °C. The residue was taken up in a solution
Hz, 2H), 7.13 (dd, J = 8.5 and 2.3 Hz, 2H), 7.06 (d, J = 2.1 Hz, 2H),
3.76 (t, J = 7.7 Hz, 2H), 2.28 (s, 6H), 2.04 (m, J = 7.4 Hz, 2H), 0.9 (t,
of 240 g of NaOH in 500 mL of water. This was added to a mixture of
600 mL of concentrated HCl and ice. This mixture was extracted with
J = 7.4 Hz, 3H). 13C NMR (CDCl ): 149.68, 142.84, 131.07, 126.35,
3
1
.5 L of ether. The extract was washed with water and then dried
126.05, 114.23, 108.76, 51.23, 28.07, 15.08, 12.17. GC-MS (EI) m/z:
(
MgSO ) and concentrated under reduced pressure to give 526.95 g of
306.1 [M] Anal. Calcd for C H N O : C, 74.49; H, 5.92; N, 9.14.
4
19 18
2
2
a brown glass (100%). This was determined to be 10% of the
ortho-,ortho-isomer, 47% of the ortho-,para-isomer, 36% of the
para-,para- isomer, and about 6% trimers. Three successive
recrystallizations from hot toluene provided pure 4,4′-(propane-1,1-
diyl)bis(3-methylphenol) (5) in an overall yield of 9%. The filtrate
from the recrystallizations was allowed to stand for several days upon
which crystals of 2-(1-(4-hydroxy-2-methylphenyl)propyl)-5-methyl-
phenol (7) separated from the mixture and were collected. Bisphenol
Found: C, 74.22; H, 6.13; N. 9.07.
2-Cyanato-1-(1-(4-cyanato-2-methylphenyl)propyl)-4-methyl-
benzene: (3). A mixture of 7 (15.05 g, 59 mmol) and cyanogen
bromide (13.06 g, 123 mmol, 2.1 equiv) in 250 mL of THF was
cooled to −30 °C, and triethylamine (18 mL, 129 mmol, 2.2 equiv)
was added dropwise. The mixture was allowed to warm to ambient
temperature, stirred for 30 min, and then washed with 250 mL of
brine. The organic layer was dried with MgSO and then concentrated
under vacuum to give 18.75 g of an orange oil. This was
4
31
5
was spectroscopically identical to a previous report. Bisphenol 7
1
was obtained as a white powder; mp 120−125 °C. H NMR (acetone-
chromatographed on silica gel using 15% EtOAc/hexanes to give
1
d ): 8.1 (s, 1H), 7.9 (s, 1H), 7.12 (d, 1H), 6.86 (d, 1H), 6.61 (m, 4H),
10.01 g of a clear liquid (55%). H NMR (acetone-d
6
): 7.15−7.43 (m,
6
.35 (t, 1H), 2.19 (s, 3H), 2.16 (s, 3H), 1.9 (m, 2H), 0.88 (t, 3H). 13
C
13
4
6H), 4.38 (t, 1H), 2.37 (d, 6H), 2.04 (q, 2H), 0.92 (t, 3H). C NMR
(acetone-d ): 151.22, 151.10, 140.62, 139.76, 139.20, 129.33, 129.16,
NMR (acetone-d ): 154.96, 154.45, 137.74, 135.79, 134.03, 129.01,
6
6
1
27.83, 127.56, 120.15, 116.87, 115.44, 112.23, 39.53, 20.08, 18.94,
129.05, 127.96, 116.79, 114.95, 112.61, 108.69, 108.64, 40.11, 28.10,
1
2.11.
20.04, 18.87, 11.71. GC-MS (EI) m/z: 306.1 [M]. Anal. Calcd for
4
,4′-(Propane-1,1-diyl)bis(1-cyanato-3-methylbenzene) (1). A
C H N O : C, 74.49; H, 5.92; N, 9.14. Found: C, 74.46; H, 6.03; N.
9.09.
19 18 2 2
mixture of 5 (23.17 g, 90.4 mmol) and cyanogen bromide (20.11 g,
0
.190 mol, 2.1 equiv) in 300 mL of THF was cooled to −30 °C while
Preparation of Networks. Prior to cure reactions, monomers
were further purified by dissolution in dichloromethane, followed by
passage through a W-Prep2XY Yamazen flash chromatography
column, and removal of the solvent under reduced pressure. All
monomers were first melted at 90−120 °C. The monomer was then
either placed directly into a hermetically sealed differential scanning
calorimetry (DSC) pan or degassed under 300 mmHg and poured into
one or more silicone rubber molds having a disc-shaped cavity and
cured using a predetermined time−temperature schedule under
flowing nitrogen, then cooled, and demolded. The resultant discs
measured approximately 12 mm in diameter by 4 mm thick.
triethylamine (28 mL, 0.201 mol, 2.2 equiv) was added dropwise. The
mixture was allowed to warm to room temperature and stirred for 30
min then washed with 300 mL of brine, dried with MgSO , and
concentrated under reduced pressure to give 32.89 g of a tan liquid.
This was chromatographed on silica gel using 15% ethyl acetate/
hexanes to give 22.78 g of a clear liquid (82%). A single peak was
observed in the GC/MS chromatogram, and the NMR data was
4
31
identical to that from a previous report.
,1-Bis(4-hydroxy-3-methylphenyl)propane (6). A round-bottom
flask equipped with magnetic stir bar was charged with o-cresol (20 g,
85 mmol, 2 equiv), H O (40 mL), and propionaldehyde (5.2 g, 90
1
1
Characterization Techniques for Networks. Differential
Scanning Calorimetry (DSC). 5−10 mg pieces of cured networks
were removed from a molded disc and hermetically sealed in an
aluminum DSC pan. Samples were then ramped under 50 mL/min of
flowing nitrogen at 10 °C/min, first heating to 350 °C, cooling to 100
°C, and then reheating to 350 °C, using a TA Instruments Q200
differential scanning calorimeter. Residual enthalpies of cure were
determined based on an algorithm for generating baselines described
2
mmol). The mixture was cooled in an ice bath while a 48% aqueous
HBr solution (40 mL, 237 mmol) was added dropwise over 1.5 h.
Once the addition was complete, the cooling bath was removed, and
the mixture was stirred at room temperature for 48 h. Afterward, the
mixture was partitioned between Et O (300 mL) and H O (300 mL).
The organic layer was washed again with H O (300 mL) and then
brine (300 mL), dried over MgSO , and filtered. The solvent was
removed under reduced pressure, and the remaining oil was
fractionally distilled at 1 Torr. The excess o-cresol distilled first,
followed by the bisphenol as a thick, pale yellow oil. The crude
bisphenol was recrystallized from cyclohexane to give the product as
2
2
2
4
3
2
in a previous publication.
Thermogravimetric Analysis (TGA). Pieces of cured discs weighing
approximately 5 mg were removed and placed in a TA Instruments
Q5000 thermogravimetric analyzer. These samples were then heated
to 600 °C, under nitrogen, and in separate experiments, in air, at 10
°C/min.
34
1
colorless needles; mp 85−87 °C [lit. 94 °C]. H NMR (DMSO-d ):
6
8
2
1
.97 (s, OH), 6.89 (d, J = 1.7 Hz, 2H), 6.84 (dd, J = 8.4 and 2.3 Hz,
H), 6.65 (d, J = 8.2 Hz, 2H), 3.47 (t, J = 7.7 Hz, 1H), 2.06 (s, 6H),
.88 (pent, J = 7.1 Hz, 2H), 0.77 (t, J = 7.1 Hz, 3H). 13C NMR
Oscillatory Thermomechanical Analysis (OTMA). Cured discs
were also tested via oscillatory thermomechanical analysis (OTMA)
with a TA Instruments Q400 series analyzer under 50 mL/min of
nitrogen flow. The discs were initially held in place with a compressive
force of 0.2 N using the standard ∼5 mm diameter flat cylindrical
probe. The force was then modulated at 0.05 Hz over an amplitude of
0.1 N (with a mean force of 0.1 N), and the temperature was ramped
twice (heating and cooling) between −50 and 200 °C (to aid in
determination of thermal lag) followed by heating to 350 °C, cooling
to 100 °C, and reheating to 350 °C, all at 50 °C/min. For samples
previously exposed to hot water, the heating rate was decreased to 20
°C/min. The order of segments was as follows: heating to 350 °C,
cooling to 100 °C, two cycles between 100 and 200 °C for thermal lag
determination, and finally heating to 350 °C. The details of the
(
DMSO-d ): 153.24, 136.06, 129.66, 125.32, 123.27, 114.31, 51.03,
6
2
8.25, 16.09, 12.81. Anal. Calcd for C H O : C, 79.65; H, 7.86.
17 20 2
Found: C, 78.90 ; H, 7.87.
1
,1-Bis(4-cyanato-3-methylphenyl)propane (2). 6 (1.35 g, 5.27
mmol) was dissolved in acetone (15 mL). The mixture was then
cooled and stirred in a −20 °C bath. Next, cyanogen bromide (1.6 g,
15.8 mmol, 3 equiv) was added to the mixture in one portion and
allowed to dissolve completely. Triethylamine (1.2 g, 11.9 mmol, 2.2
equiv) was then added dropwise over 30 min with the aid of an
addition funnel. Shortly after the addition was started, a white
precipitate formed. Afterward, the mixture was stirred at ambient
temperature for 1.5 h. The reaction was partitioned between Et O (25
mL) and H O (50 mL). The organic layer was separated and washed
2
32
thermal lag determination procedure have been reported previously.
2
with H O (2 × 50 mL) and then with brine (50 mL). The organic
Infrared Spectroscopy. The cured discs were cut using a Struers
rotary saw with a diamond blade to expose a fresh, smooth surface.
Fourier transform infrared spectroscopy (FTIR) was carried out on
these surfaces using a Thermo Scientific Nicolet 6700 spectrometer in
attenuated total reflectance mode with the ZnSe crystal attachment. A
total of 512 scans were completed on these surfaces with a resolution
2
layer was separated, dried over MgSO , and filtered, and the solvent
4
was then removed under reduced pressure, leaving a colorless oil. The
oil was stored under vacuum and eventually began to crystallize. The
crude solid was slurried with isopropyl alcohol (15 mL) and filtered on
a medium porosity glass fritted filter funnel. Further purification by
−1
recrystallization from heptane gave the product as colorless needles
of 4 cm to obtain spectra. Background values at 1850, 2700, and
1
−1
(
840 mg, 50%); mp 95−97 °C. H NMR (CDCl ): 7.35 (d, J = 8.5
3300 cm were used to fit a quadratic global baseline. After
3
C
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Table 1. Summary of DSC Data
monomer
TG,monomer (°C)
Tm (°C)
n/a
ΔHm (kJ/mol)
ΔHcure (J/g)
ΔHcure (kJ/equiv)
Tcure‑onset (°C)
Tpeak‑cure (°C)
1
2
3
−35 ± 1
n/a
n/a
714 ± 58
717
109 ± 9
110
276 ± 8
271
302 ± 3
303
99.1
39
−22 ± 1
62.6 ± 0.3
24 ± 1
710 ± 49
109 ± 8
298 ± 12
318 ± 6
subtracting the global baseline, the integrated area between 1480 and
−1
−1
1
510 cm , containing the aromatic peak near 1500 cm , was used as
a reference for normalization. After subtracting the global baseline, the
−1
integrated area between 2200 and 2300 cm was calculated, and
subtracted from a local secondary baseline between values at 2450 and
−1
2
150 cm , to determine the peak area associated with residual cyanate
ester groups.
Other Physical Characterization. Sample densities were deter-
mined using a Mettler-Toledo analytical balance with an attachment
for specific gravity determination. Ethanol (ACS Reagent grade from
Sigma-Aldrich) was used as the immersion medium, as it is superior to
water for avoiding formation of bubbles adhered to the samples. A
minimum of four weight measurements were collected and recorded
per sample, making use of the cured discs prior to removal of portions
for any other testing. For water immersion testing, cured discs were
dried to a ± 0.0001 g constant weight in a vacuum desiccator, then
weighed, and immersed in approximately 250 mL of deionized water
maintained at 85 °C for 96 h. After removal from the water, samples
were patted dry and weighed to determine the moisture uptake (on a
dry weight basis).
Figure 4. DSC (first heating) of 3 after purification by flash
chromatography, compared to newly synthesized sample of 1 for
reference.
exotherm for 3 occurs at higher temperatures than for 1,
possibly indicating that a lower level of catalytic impurities
35
(
such as phenol-containing compounds) was obtained during
RESULTS AND DISCUSSION
Monomer and Cure Characteristics. Note that a new
batch of 1 was synthesized from m-cresol for use in this study;
chromatography of 3. The impurity profile of 3 is important to
consider during subsequent comparisons of the properties of
networks derived from this compound.
Figure 5 shows the second heating scans of these (now
cured) monomers. A very underappreciated aspect of the
■
3
1
its properties match those previously reported (see
Supporting Information). Isomers 2 and 3 of 1 exhibit behavior
typical of highly pure cyanate ester monomers. Key features of
DSC data are listed in Table 1. Figure 3 shows the DSC trace
Figure 5. DSC (second heating) of 2 and 3 after purification by flash
chromatography, compared to newly synthesized sample of 1 for
reference. Note that the initial DSC heating results in cure
temperatures in excess of 300 °C, which leads to some degradation
Figure 3. DSC (first heating) of 2 after purification by flash
chromatography, compared to newly synthesized sample of non-
crystalline 1 (after similar purification) for reference.
and a T below the maximum possible value for the cured networks.
G
commonly performed analysis of DSC second heating scans for
high-temperature thermosetting networks is that the vast
majority of the cure that takes place during the first DSC
scan involves temperatures (well over 250 °C in this case) that
are significantly higher than those typically utilized to cure the
networks without degradation (typically 150−200 °C for
cyanate esters). Under these conditions, side reactions that
destroy the reactive groups prior to network formation may be
significant, leading to TG values that are well below those
achieved by completely curing the network more slowly at
lower temperatures. Thus, a reliable inference from Figure 5 is
that under typical cure conditions for polycyanurates, relatively
for 2 (uncured) compared to 1. Although 1 is an extremely
stable supercooled liquid, 2 crystallizes readily and thus exhibits
a melting point just below 100 °C in the DSC trace. The cure
characteristics of both are very similar, indicating that a similar
level of monomer purity was attained. Because DSC cure
characteristics are highly dependent on monomer purity, the
effect of the chemical structural differences between 1 and 2 on
cure characteristics is not reliably attainable from an analysis of
the DSC data.
Figure 4 provides a DSC comparison on uncured 1 and 3.
Although a melting point for 3 was observed in some sample
runs, a stable supercooled liquid was readily attainable under
quiescent conditions in the DSC instrument. The cure
stable networks with a T of over 200 °C may be formed from
G
these monomers.
D
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Network Characteristics. Conversion data for the net-
works after cure were derived from a maximum-likelihood
tion) were carried out with the aid of the correlations
developed by Bicerano and Georjon and Galy. The results,
shown in Figure 7, indicate that (despite some anomalous
41
38
statistical algorithm that combines DSC residual cure and T
G
36
data, FTIR data, and the diBenedetto equation. The method
is detailed in the Supporting Information and enables
measurement of conversions typically to a precision of better
than 1%. This level of precision is important in establishing the
structure−property relationships that are discussed below.
The weight gain data resulting from exposure to deionized
water at 85 °C for 96 h for the cured networks is shown in
Figure 6. Remarkably, despite the chemical similarity and nearly
Figure 7. Packing fraction at 293 K of cured networks.
individual data points) the packing fraction is highest (and
therefore free volume is lowest) in network 1, with slightly
more free volume in network 2 and considerably more in
37,38
network 3. As is typical for polycyanurate networks,
the
free volume increases steadily as conversion increases, across
the entire range of conversions studied.
Figure 6. Water uptake after immersion for 96 h at 85 °C of cured
networks.
Network 1, with the highest packing fraction and lowest free
volume, exhibits the highest water uptake. In fact, as seen with
comparisons among other polycyanurate networks of varied
chemical composition, there is no apparent correlation between
overall free volume (at a given conversion) and moisture
identical density of favorable sites for water uptake in these
networks (see Supporting Information for calculations), the
observed weight gains differ by up to a factor of 4. Thus, the
extent of water uptake appears to depend very substantially on
the detailed physical structure of the network segments in the
vicinity of these sites. In accord with previous studies on the
effect of methylation ortho to cyanate ester groups on water
42
uptake. Moisture uptake characteristics therefore appear to
depend far more on local details of the structure near preferred
sites for water uptake than on overall network physical
properties. Nonetheless, as with previous comparative data
among networks with different monomer chemical composi-
3
2,33
32,37,43
uptake,
the presence of the methyl groups in network 2
tion,
networks with lower glass transition temperatures at
results in a roughly 50% decrease in water uptake compared to
network 1, at all conversions examined. This effect has been
attributed to steric hindrance of the cyanurate oxygen, which is
a favored site for water uptake due to its polarity. Electronic
effects caused by the ortho substitution of an aliphatic group
near the cyanurate oxygen also may play some role.
full cure exhibit characteristically lower moisture uptake.
Whereas previous speculation about the mechanism underlying
this correlation focused on differences in the density of favored
sites for water uptake, in the current work the density of these
sites changes by less than 2% (see Supporting Information) at a
given conversion.
42
Perhaps even more remarkably, network 3, in which only one
of the two cyanate ester groups has an ortho substituent, namely
the bridge rather than a methyl group, exhibits nearly the same
comparative decrease in water uptake as network 2, in which
both cyanate ester groups are ortho to methyl groups. Only
when the effect of conversion on water uptake is considered
does network 2 provide a clearly superior characteristic. In
In addition to slowing down hydrolysis of the networks by
lowering the concentration of water, the local environment
44
around the hydrolyzable aryl ether bonds in networks 2 and 3
may also inhibit hydrolysis. As a result, the decrease in glass
transition temperature on exposure to water at elevated
temperatures, a major concern for applications of polycyanurate
networks, is mitigated as shown in Figure 8, which illustrates
32,37−40
agreement with numerous previous studies,
the water
the measured “wet” T values measured by dynamic TMA (raw
G
uptake increases with increasing conversion above about 85%
conversion but, where data are available, shows a decrease in
uptake with increasing conversion below about 85% conversion.
The former effect is suppressed in network 2, as in previously
studied ortho-methylated cyanate ester networks, but only
partly suppressed in network 3. These differences reinforce the
conclusion that moisture uptake is highly dependent on the
physical and chemical environment around favored sites for
water uptake.
data in Supporting Information) compared to the dry T values
G
36
predicted from the diBenedetto equation. These dry TG
values are generated from the combined DSC/FTIR analysis
for conversions described in the Supporting Information. It is
important to note that in situ cure of the “wet” samples during
3
2
14
analysis also takes place, so that the “wet” network conversion
during measurement (which is difficult to determine) is higher
than the initial conversion of the as-prepared dry networks used
to construct Figure 8.
A simple way of examining the characteristics of the networks
in order to gain insight into how physical structure impacts
properties such as moisture uptake is to calculate the packing
fraction. These calculations (provided in Supporting Informa-
Figure 8 clearly shows that while the “wet” T values for
G
network 1 are lower than the dry values across the range of
conversions studied, those for network 2 are nearly the same or
slightly higher, while those for network 3 are higher at low
E
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 9. “Wet, fully-cured” glass transition temperature (after 96 h
immersion in water at 85 °C, followed by heating to 350 °C at 20 °C/
min).
Figure 8. “Wet” glass transition temperature (after 96 h immersion in
water at 85 °C). The diBenedetto equation fits of “dry” glass transition
temperatures measured by DSC as a function of initial network
conversion are indicated by the lines.
lower still for 2, matching the pattern seen for samples exposed
to water but not heated to 350 °C, indicating the water
exposure is likely the main driver of the loss in cross-link
density. Interestingly, networks with a lower initial conversion
prior to exposure show the least loss in cross-link density. If
carbamate formation were the main driver of loss in cross-link
density, one would expect the opposite trend. Thus, cyanurate
hydrolysis appears to be the main driver of loss in cross-link
density.
conversions, but lower at high conversions. Comparison with
Figure 6 shows that whenever water uptake is low and
decreases slowly with increasing conversion, there is little
hydrolysis, and the “wet” T tends to be a bit higher than the
G
dry value due to in situ cure as discussed earlier. In contrast,
whenever the water uptake increases significantly with
increasing conversion, hydrolysis lowers the T of the “wet”
G
network to a significant extent. A large increase in water uptake
One way to compare the extent of hydrolysis and carbamate
with increasing conversion is thought to signify a relatively
formation in each network is to subtract the dry T measured
G
“
open” network structure in which cyanurates are accessible to
in a DSC scan of an identical sample heated to 350 °C from the
measured (by dynamic TMA) wet TG of each sample after
heating to 350 °C. This method effectively baselines each
3
2
water, so it seems logical that under such circumstances
hydrolysis would be enhanced. For applications, however, the
absolute value of the “wet” T is often more important than the
network to its maximum attainable T value taking into account
G
G
rate of hydrolysis, depending on exposure conditions. For the
conditions tested, network 3 at about 95% conversion performs
thermochemical degradation during heating. This method is
not perfect because sample sizes, thermal lag, and heating rates
all differ between the two methods, so that the exposure
histories are not truly identical except for the presence of water.
Nonetheless, the results are quite informative, as shown in
Figure 10.
best because it combines a relatively high initial dry T with a
G
more limited extent of hydrolysis. The higher dry T values
G
may be attributed to the relatively bulky substitution around the
bridge between aromatic rings in networks 1 and 3. By
combining the bulky substitution pattern that increases TG
values generally with a local environment that inhibits water
uptake around the cyanurate oxygen, network 3 provides an
optimized structure for high “wet” T values over a range of
G
initial conversions from around 70−95%.
In addition to hydrolysis leading to the loss of cross-linking,
water may also affect polycyanurate networks by converting
reactive cyanate ester groups into carbamate groups, which then
4
5
decompose with loss of CO on heating above about 180 °C,
2
with the net effect being the conversion of cross-linkable chain
ends into “loose” ends. Such a transformation lowers the T of
G
the network at full cure. Measurements of the T of networks at
G
full cure after exposure to water, accomplished by heating the
Figure 10. Difference between “wet” and “dry” glass transition
temperature of fully cured single-component networks. This difference
represents a relative measure of the amount of degradation induced by
exposure to water at elevated temperature.
“
wet” networks to 350 °C, thus reveal information about the
extent of both cross-link loss by hydrolysis (which also lowers
the T ) and carbamate formation. A complicating factor is that
G
heating of the network to 350 °C even under dry conditions
46
can lead to some degradation and concomitant loss in TG.
Figure 9 shows the measured T values for the “wet” samples
According to Figure 10, network 1 experiences the greatest
G
after heating to 350 °C. From the DSC/FTIR analysis, the
loss in T due to exposure to water at elevated temperatures
G
expected T values for fully cured, nondegraded networks are
(20−40 °C), while networks 2 and 3 experience almost no loss
and, in some cases, a net gain due to in situ cure of the “wet”
samples during measurement. Extrapolating the trends to
complete conversion eliminates both in situ cure effects as well
as any differences in carbamate generation, resulting in one of
the best comparative measures of the effect of network
G
2
74 °C for 1, 244 °C for 2, and 266 °C for 3. None of the
networks exhibit T values close to those expected for intact,
G
fully cured networks, indicating that water exposure followed by
heating to 350 °C induces significant loss of cross-link density.
The T values are highest for 3, slightly lower for 1, and slightly
G
F
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
structure on hydrolysis rate. By this measure, networks 2 and 3
are not significantly different; both are significantly better than
network 1 at preventing hydrolysis despite the different local
environments around the cyanurate oxygen atoms. Note also
that the dependence of T loss on conversion is only about half
G
as great as that seen in Figure 9, indicating that even “dry”
samples with higher initial conversions end up with lower “fully
cured” T values. The most likely explanation for this effect is
G
that the higher cure temperatures required to generate the
higher conversions lead to increased side reactions that destroy
the reactive cyanate ester groups during cure. A more detailed
analysis of these effects is provided in the Supporting
Information.
Because the local environment that inhibits hydrolysis
reactions is also present in the monomer, a potential negative
consequence is increased difficulty in achieving complete
cyclotrimerization of the networks. Once vitrification has
taken place, it could be argued that issues such as steric
hindrance at reactive sites take on less importance compared to
the overall dynamics of diffusion within the network. One way
to investigate these issues is to examine the observed network
TG as a function of cure temperature. Figure 11 shows
Figure 12. Extent of network conversion compared to the expected
vitrification point (a measure of the ease of cure). The dotted line
represents the expected result if cure were to cease completely upon
vitrification, which is generally not the case for polycyanurates. Points
further above and to the left of the line indicate a greater extent of cure
after vitrification.
from lower glass transition temperatures at high conversions,
which enabled these networks to avoid vitrification during cure.
Methylation ortho to cyanate ester groups may therefore
facilitate cure by lowering glass transition temperatures
(
avoiding vitrification), even though once vitrification takes
place, these groups may hinder further cure. Of additional note,
networks 2 and 3 achieved complete conversion at the highest
cure temperature studied (230 °C), while network 1 did not
(
see the conversion data in the Supporting Information). This
result may indicate that degradation during cure at elevated
temperature (also discussed in the Supporting Information) is
more pronounced in network 1, while decreased degradation in
networks 2 and 3 facilitates complete cure by enabling higher
cure temperatures to be utilized without degradation.
A final stability consideration that is worthy of examination is
the thermochemical stability of the cured network. Figure 13
Figure 11. Observed dry glass transition temperature as a function of
cure temperature (all samples cured for 24 h). The dotted line
indicates the equivalence of the cure and glass transition temperatures.
For data points above and to the left of this line, vitrification takes
place during cure.
comparative TG data for the networks used in this study.
Network 1 consistently achieved the highest T values, even
G
when (as at 170 °C) networks 2 and 3 did not vitrify during
cure. The T values achieved for networks 2 and 3 were quite
G
similar, despite the differences in monomer structure.
One shortcoming of Figure 11 is that it does not take into
account different diBenedetto equation parameters for different
networks. In particular, at high cure temperatures the T for
G
each network will approach a separate maximum attainable
value, while the conversion at which vitrification takes place is
different for different networks cured at the same temperature.
In Figure 12, the conversions achieved during cure are plotted
as a function of the conversion at vitrification. Because
conversion represents a “universal” scale for all networks,
comparisons among networks are facilitated.
Figure 13. TGA mass loss traces in air. Note that all networks were
cured at 230 °C for 24 h prior to the start of the TGA experiment.
Using this representation, network 3 shows behavior that is
intermediate between networks 1 and 2. Network 1
consistently achieved conversions well beyond the vitrification
shows TGA traces in air for each network. (A tabular summary
of TGA data is provided in the Supporting Information.)
Interestingly, networks 1 and 3 exhibit remarkably similar
behavior, while network 2 appears to be significantly more
stable. Figure 14 shows analogous TGA data under nitrogen,
with a separately synthesized batch of network 2 added in order
to assess the effect of variable trace impurities found during
47,48
point (a common occurrence for polycyanurate networks)
while network 2 failed to do so under identical cure and
conversion conditions. These data indicate that the relatively
higher conversions achieved in network 2 may result principally
G
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
atoms, whether provided by methyl groups or aliphatic bridge
groups, significantly lowered the moisture uptake of the
networks and led to decreased loss in glass transition
temperatures due to hydrolysis. In terms of absolute “wet”
glass transition temperatures, however, the properties of
ortho,para-linked aryl cyanates proved superior to ortho-
methylated polycyanurate due to the fact that the former
provides a higher dry glass transition temperatures of cured
networks compared to the latter. Both forms of ortho-
substitution still enabled the production of networks with
>
99% conversion at cure temperatures of 230 °C. Placement of
methyl groups in an ortho-position, however, appeared to offer
some benefits compared to placement of groups in the meta-
position, even in systems with ortho-substitution provided by
the bridge, with respect to thermochemical stability.
Figure 14. TGA mass loss traces under nitrogen, with a separately
synthesized batch of network 2 added to assess reproducibility. Note
that all networks were cured at 230 °C for 24 h prior to the start of the
TGA experiment.
ASSOCIATED CONTENT
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.macro-
mol.7b00824.
■
*
S
synthesis. Although the char yields do not change much
between batches, the initial decomposition temperature (likely
NMR data (S1); comparison of properties of 1 with
previously published data (S2); determination of
conversions for networks, including DSC and FTIR
data (S3); calculation of van der Waals volumes (S4);
determination of key number density parameters for
networks (S5); dynamic TMA data for “wet” samples
49
a reflection of decomposition kinetics) does show substantial
variation, indicating that as previously reported, impurities play
50
a key role in thermochemical stability. (The other physical
properties and chemical analysis of the two batches of 2 were
very similar, and the observed difference in the onset of
degradation between the two batches of about 30 °C is many
times larger than the estimated measurement error of 3−5 °C
(
S6); examination of “fully cured” glass transition
temperature data (S7); tabular summary of TGA data
S8); GC-MS parameters (S9) (PDF)
42
for this technique. )
(
If variability due to impurities is a significant factor in the
thermochemical decomposition kinetics of polycyanurate
networks, then the remarkable similarity in the decomposition
behavior of networks 1 and 3 requires an explanation. The
most interesting potential explanation is that because 1 and 3
are synthesized by the same reaction technique with the same
starting materials, using only different separation procedures to
isolate one product or the other, the same set of impurities is
generated. Another potential explanation is that the ortho-
methylation present in 2 inhibits a degradation pathway by a
mechanism that does not involve steric effects, such as
electronic induction. If the inhibited mechanism, which may
even be accelerated in a meta-methylated network, is a
characteristic of the network segments themselves, rather
than impurities, it may be the rate-determining step, leading to
consistent behavior. Only in its absence would mechanisms
involving impurities become rate-determining. Examining the
char yields in nitrogen, which are insensitive to kinetics, does
show that network 2 performs consistently better than 1 or 3,
so that methylation ortho to cyanurate groups does lead to
improved thermochemical stability, in at least one aspect,
compared to networks with substitution ortho to the bridge
linkage.
AUTHOR INFORMATION
Corresponding Author
E-mail andrew.guenthner@us.af.mil (A.J.G.).
■
*
ORCID
Andrew J. Guenthner: 0000-0002-6640-0176
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
The support of the Office of Naval Research and the Air Force
Office of Scientific Research is gratefully acknowledged.
■
REFERENCES
■
(
1) Chemistry and Technology of Cyanate Ester Resins, Hamerton, I.,
Ed.; Chapman & Hall: London, 1994.
2) Fang, T.; Shimp, D. A. Polycyanate Esters – Science and
Applications. Prog. Polym. Sci. 1995, 20, 61−118.
3) Nair, C. P. R.; Mathew, D.; Ninan, K. N. In New Polymerization
(
(
Techniques and Synthetic Methodologies; Abe, A., Albertsson, A.-C.,
Cantow, H. J., Eds.; Springer-Verlag: Berlin, 2001; Vol. 155, pp 1−99.
(4) Hamerton, I.; Hay, J. N. Recent technological developments in
cyanate ester resins. High Perform. Polym. 1998, 10, 163−174.
(5) Arepalli, S.; Moloney, P. Engineered nanomaterials in aerospace.
MRS Bull. 2015, 40, 804−811.
CONCLUSIONS
■
The configuration of network segments, primarily through
various patterns of substitution on the phenyl rings, particularly
near cyanurate oxygen atoms, has a substantial effect on the
physical properties, cure properties, and stability of poly-
cyanurate networks. These effects were illustrated via synthesis,
network formation, and characterization of an isomeric series of
n-propyl-bridged cyanate ester monomers. Substitution of
aliphatic groups in a position ortho- to cyanurate oxygen
(
6) Dawson, D. Composites carry the Curiosity rover to a safe Mars
landing High Performance Composites, May 2014; http://www.
compositesworld.com/articles/composites-carry-the-curiosity-rover-
to-a-safe-mars-landing, accessed 20 Nov 2016.
(
7) Hill, J.; Munshi, N.; Tupper, M.; Babcock, H.; Kintz, A.; Kearns,
T.; Kimoto, A. Risk Mitigation for Cyanate Ester Insulation of Large
Magnets Through Cure Optimization. IEEE Trans. Appl. Supercond.
2015, 25, 1−5.
H
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
(
8) Humer, K.; Prokopec, R.; Weber, H. W.; Fillunger, H.; Maix, R.
(25) Puchot, L.; Verge, P.; Fouquet, T.; Vancaeyzeele, C.; Vidal, F.;
Habibi, Y. Breaking the symmetry of dibenzoxazines: a paradigm to
tailor the design of bio-based thermosets. Green Chem. 2016, 18,
3346−3353.
K. Characterization and Qualification of Advanced Insulators for
Fusion Magnets. Fusion Eng. Des. 2013, 88, 350−360.
(
9) Bittner-Rohrhofer, K.; Humer, K.; Fillunger, H.; Maix, R. K.;
Weber, H. W.; Fabian, P. E.; Munshi, N. A. Radiation Effects on the
Mechanical Integrity of Novel Organic Insulators for the ITER Magnet
Coils. J. Nucl. Mater. 2004, 329−333, 1083−1087.
(26) Hannoda, Y.; Akasaka, Y.; Shibata, M. Bio-based thermosetting
bismaleimide resins using cardany linolenate and allyl cardanyl ether.
React. Funct. Polym. 2015, 97, 96−104.
(
10) Idesaki, A.; Nakamoto, T.; Yoshida, M.; Shimada, A.; Iio, M.;
(27) Shibata, M.; Shimasaki, T.; Satoh, H.; Iwai, M.; Neda, M. Bio-
Sasaki, K.; Sugano, M.; Makida, Y.; Ogitsu, T. Development of high
radiation-resistant glass fiber reinforced plastics with cyanate-based
resin for superconducting magnet systems. Fusion Eng. Des. 2016, 112,
based thermosetting resins composed of aliphatic polyol-derived
polymaleimides and allyleugenol. React. Funct. Polym. 2015, 97, 69−76.
(28) Shibata, M.; Tetramoto, N.; Imada, A.; Neda, M.; Sugimoto, S.
4
(
18−424.
Bio-based thermosetting bismaleimide resins using eugenol, bieugenol
11) Fyfe, C. A.; Niu, J.; Rettig, S. J.; Burlinson, N. E. High-
and eugenol novolac. React. Funct. Polym. 2013, 73, 1086−1095.
Resolution Carbon-13 and Nitrogen-15 NMR Investigations of the
Mechanism of the Curing Reactions of Cyanate-Based Polymer Resins
in Solution and the Solid State. Macromolecules 1992, 25, 6289−6301.
(29) Shibata, M.; Teramoto, N.; Nakamura, Y. High performance
bio-based thermosetting resins composed of tung oil and bismalei-
mide. J. Appl. Polym. Sci. 2011, 119, 896−901.
(
12) Guenthner, A. J.; Reams, J. T.; Ford, M. D.; Lamison, K. R.;
(30) Hirayama, K.-I.; Irie, T.; Teramoto, N.; Shibata, M. High-
Mabry, J. M. In Effect of in-Situ Cure on the Measurement of Glass
Transition Temperatures in High-Temperature Thermosetting Polymers;
Society for the Advancement of Material and Process Engineering:
performance bio-based thermosetting resins composed of dehydrated
castor oil and bismaleimide. J. Appl. Polym. Sci. 2009, 114, 1033−1039.
(31) Harvey, B. G.; Guenthner, A. J.; Lai, W. W.; Meylemans, H. A.;
2015; SKU A-8568, 16 pp.
Davis, M. C.; Cambrea, L. R.; Reams, J. T.; Lamison, K. R. Effects of o-
Methoxy Groups on the Properties and Thermal Stability of
Renewable High-Temperature Cyanate Ester Resins. Macromolecules
(
13) Davis, M. C.; Guenthner, A. J.; Groshens, T. J.; Reams, J. T.;
Mabry, J. M. Polycyanurate networks from anethole dimers: Synthesis
and characterization. J. Polym. Sci., Part A: Polym. Chem. 2012, 50,
2
(
015, 48, 3173−3179.
4
(
127−4136.
32) Guenthner, A. J.; Wright, M. E.; Chafin, A. P.; Reams, J. T.;
14) Davis, M. C.; Guenthner, A. J.; Sahagun, C. M.; Lamison, K. R.;
Lamison, K. R.; Ford, M. D.; Kirby, S. P. J.; Zavala, J. J.; Mabry, J. M.
Mechanisms of Decreased Moisture Uptake in ortho-Methylated
Di(Cyanate Ester) Networks. Macromolecules 2014, 47, 7691−7700.
Reams, J. T.; Mabry, J. M. Polycyanurate networks from dehydroanet-
hole cyclotrimers: Synthesis and characterization. Polymer 2013, 54,
6
(
902−6909.
(33) Shimp, D. A. The translation of dicyanate structure and
15) Cash, J. J.; Davis, M. C.; Ford, M. D.; Groshens, T. J.;
Guenthner, A. J.; Harvey, B. G.; Lamison, K. R.; Mabry, J. M.;
Meylemans, H. A.; Reams, J. T.; Sahagun, C. M. High Tg
thermosetting resin from resveratrol. Polym. Chem. 2013, 4, 3859−
cyclotrimerization efficiency to polycyanurate properties. Polym. Mater.
Sci. Eng. 1986, 54, 107−113.
(34) Harden, W. C.; Reid, E. E. Condensation of certain phenols with
some aliphatic anhydrides. J. Am. Chem. Soc. 1932, 54, 4325−4334.
3865.
(
16) Cambrea, L. R.; Davis, M. C.; Garrison, M. D.; Groshens, T. J.;
́
(35) Gomez, C. M.; Recalde, I. B.; Mondragon, M. Kinetic
Lyon, R. E.; Safronava, N. Processable cyanate ester resin from Cis
resveratrol. J. Polym. Sci., Part A: Polym. Chem. 2017, 55, 971−980.
parameters of a cyanate ester resin catalyzed with different proportions
of nonylphenol and cobalt acetylacetonate catalyst. Eur. Polym. J. 2005,
41, 2734−2741.
(
17) Harvey, B. G.; Sahagun, C. M.; Guenthner, A. J.; Groshens, T. J.;
Cambrea, L. R.; Reams, J. T.; Mabry, J. M. A High-Performance
Renewable Thermosetting Resin Derived from Eugenol. ChemSu-
sChem 2014, 7, 1964−1969.
(36) Pascault, J. P.; Williams, R. J. J. Glass transition temperature
versus conversion relationships for thermosetting polymers. J. Polym.
Sci., Part B: Polym. Phys. 1990, 28, 85−95.
(
18) Harvey, B. G.; Guenthner, A. J.; Yandek, G. R.; Cambrea, L. R.;
(37) Reams, J. T.; Guenthner, A. J.; Lamison, K. R.; Vij, V.; Lubin, L.
Meylemans, H. A.; Baldwin, L. C.; Reams, J. T. Synthesis and
characterization of a renewable cyanate ester/polycarbonate network
derived from eugenol. Polymer 2014, 55, 5073−5079.
M.; Mabry, J. M. Effect of Chemical Structure and Network Formation
on Physical Properties of Di(Cyanate Ester) Thermosets. ACS Appl.
Mater. Interfaces 2012, 4, 527−535.
(
19) Meylemans, H. A.; Harvey, B. G.; Reams, J. T.; Guenthner, A. J.;
(38) Georjon, O.; Galy, J. Effects of crosslink density on the
Cambrea, L. R.; Groshens, T. J.; Baldwin, L. C.; Garrison, M. D.;
Mabry, J. M. Synthesis, Characterization, and Cure Chemistry of
Renewable Bis(cyanate) Esters Derived from 2-Methoxy-4-Methyl-
phenol. Biomacromolecules 2013, 14, 771−780.
volumetric properties of high Tg polycyanurate networks. Con-
sequences on moisture absorption. Polymer 1998, 39, 339−345.
(39) Guenthner, A. J.; Sahagun, C. M.; Lamison, K. R.; Reams, J. T.;
Haddad, T. S.; Mabry, J. M. Effect of Nanoparticle Functionalization
on the Performance of Polycyanurate/Silica Nanocomposites. Ind. Eng.
Chem. Res. 2016, 55, 7096−7107.
(
20) Harvey, B. G.; Guenthner, A. J.; Meylemans, H. A.; Haines, S. R.
L.; Lamison, K. R.; Groshens, T. J.; Cambrea, L. R.; Davis, M. C.; Lai,
W. W. Renewable thermosetting resins and thermoplastics from
vanillin. Green Chem. 2015, 17, 1249−1258.
(40) Hay, J. N. Processing and cure schedules for cyanate ester resins.
In Chemistry and Technology of Cyanate Ester Resins; Hamerton, I., Ed.;
Chapman & Hall: London, 1994; pp 165−166.
(
21) Harvey, B. G.; Guenthner, A. J.; Koontz, T. A.; Storch, P. J.;
Reams, J. T.; Groshens, T. J. Sustainable hydrophobic thermosetting
resins and polycarbonates from turpentine. Green Chem. 2016, 18,
(41) Bicerano, J. Prediction of Polymer Properties, 3rd ed.; Marcel
Dekker, Inc.: New York, 2002; pp 66−78.
2
(
416−2423.
(42) Guenthner, A. J.; Lamison, K. R.; Vij, V.; Reams, J. T.; Yandek,
22) Laskoski, M.; Clarke, J. S.; Neal, A.; Harvey, B. G.; Ricks-
G. R.; Mabry, J. M. New Insights into Structure-Property Relationships
in Thermosetting Polymers from Studies of Cocured Polycyanurate
Networks. Macromolecules 2012, 45, 211−220.
Laskoski, H. L.; Hervey, W. J.; Daftary, M. N.; Shepherd, A. R.; Keller,
T. M. Sustainable High-Temperature Phthalonitrile Resins Derived
from Resveratrol and Dihydroresveratrol. ChemistrySelect 2016, 1,
(43) See: Simon, S. L.; Gillham, J. K. Cure kinetics of a thermosetting
3
(
423−3427.
liquid dicyanate ester monomer/high Tg polycyanurate material. J.
Appl. Polym. Sci. 1993, 47, 461−485. Note that the glass transition
temperature of RTX-366 at full cure of 190 °C is much lower than
typical polycyanurates, while the moisture uptake as listed in Chemistry
and Technology of Cyanate Ester Resins, Hamerton, I., Ed.; Chapman &
Hall: London, 1994; p 332 is also quite low at 0.6%.
23) Sharma, P.; Lochab, B.; Kumar, D.; Roy, P. K. Sustainable Bis-
benzoxazines from Cardanol and PET-Derived Terephthalamides.
ACS Sustainable Chem. Eng. 2016, 4, 1085−1093.
(
24) Mehta, B.; Watt, P.; Soucek, M. D.; Pugh, C. Moderate
Temperature Curing of Plant Oils with Bismaleimides via the Ene
Reaction. Ind. Eng. Chem. Res. 2016, 55, 11727−11735.
I
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
(
44) Kasehagen, L. J.; Haury, I.; Macosko, C. W.; Shimp, D. A.
Hydrolysis and blistering of cyanate ester networks. J. Appl. Polym. Sci.
997, 64, 107−113.
45) Shimp, D. A.; Ising, S. J. Moisture effects and their control in the
curing of polycyanate resins. Polym. Mater. Sci. Eng. 1992, 66, 504−
05.
46) Georjon, O.; Galy, J.; Pascault, J. P. Isothermal Curing of an
1
(
5
(
Uncatalyzed Dicyanate Ester Monomer: Kinetics and Modeling. J.
Appl. Polym. Sci. 1993, 49, 1441−1452.
(
47) Guenthner, A. J.; Davis, M. C.; Ford, M. D.; Reams, J. T.;
Groshens, T. J.; Baldwin, L. C.; Lubin, L. M.; Mabry, J. M.
Polycyanurate Networks with Enhanced Segmental Flexibility and
Outstanding Thermochemical Stability. Macromolecules 2012, 45,
9
(
707−9718.
48) Mondragon, I.; Solar, L.; Recalde, I. B.; Gomez, C. M. Cure
́
kinetics of a cobalt catalysed dicyanate ester monomer in air and argon
atmospheres from DSC data. Thermochim. Acta 2004, 417, 19−26.
(
49) Guenthner, A. J.; Reams, J. T.; Lamison, K. R.; Ramirez, S. M.;
Swanson, D. D.; Yandek, G. R.; Sahagun, C. M.; Davis, M. C.; Mabry,
J. M. Synergistic Physical Properties of Cocured Networks Formed
from Di- and Tricyanate Esters. ACS Appl. Mater. Interfaces 2013, 5,
8
(
772−8783.
50) Hamerton, I. Properties of unreinforced cyanate ester matrix
resins. In Chemistry and Technology of Cyanate Ester Resins; Hamerton,
I., Ed.; Chapman & Hall: London, 1994; pp 193−229.
J
DOI: 10.1021/acs.macromol.7b00824
Macromolecules XXXX, XXX, XXX−XXX