Biorenewable ROMP-based thermosetting copolymers from functionalized castor oil derivative with various cross-linking agents

Article abstract of DOI:10.1016/j.polymer.2014.09.023

A new norbornenyl-functionalized castor oil alcohol (NCA) was synthesized and ring-opening metathesis copolymerized separately with two norbornene-based cross-linking agents: dicyclopentadiene (DCPD) and a bifunctional norbornene crosslinker (CL). Isothermal differential scanning calorimetry (DSC) was used to examine the cure behavior of NCA/DCPD and NCA/CL resins, through which a reasonable cure schedule was determined. The glass transition temperature (T_g) and storage modulus (E′), characterized by dynamic mechanical analysis (DMA), increased significantly in both copolymer systems with the addition of cross-linking agents. Cross-link density of the two systems was evaluated using a modified empirical equation from the kinetic theory of rubber elasticity. Differences in tensile stress-strain behavior and thermal stability between polymerized NCA/DCPD and NCA/CL were correlated to the structural rigidity and cross-linking density resulting from the cross-linking agents.

Full text of DOI:10.1016/j.polymer.2014.09.023

Accepted Manuscript
Biorenewable ROMP-based Thermosetting Copolymers from Functionalized Castor
Oil Derivative with Various Cross-linking Agents
Rui Ding, Ying Xia, Timothy C. Mauldin, Michael R. Kessler
PII:
S0032-3861(14)00817-9
10.1016/j.polymer.2014.09.023
JPOL 17265
DOI:
Reference:
To appear in: Polymer
Received Date: 3 July 2014
Revised Date: 3 September 2014
Accepted Date: 9 September 2014
Please cite this article as: Ding R, Xia Y, Mauldin TC, Kessler MR, Biorenewable ROMP-based
Thermosetting Copolymers from Functionalized Castor Oil Derivative with Various Cross-linking Agents,
Polymer (2014), doi: 10.1016/j.polymer.2014.09.023.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical Abstract
22

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
Biorenewable ROMP-based Thermosetting Copolymers from Functionalized Castor
Oil Derivative with Various Cross-linking Agents
Rui Dinga, Ying Xiaa, Timothy C. Mauldina, Michael R. Kesslera,b,
*
aDepartment of Materials Science and Engineering, Iowa State University, Ames, USA
bSchool of Mechanical and Materials Engineering, Washington State University, Pullman, USA
*Corresponding author. School of Mechanical and Materials Engineering, Washington State University,
Spokane Street, Sloan Hall 201, PO Box 642920, Pullman, WA 99164-2920, USA. Tel.: +1
5093358654. E-mail address: MichaelR.Kessler@wsu.edu (M.R. Kessler).
9
ABSTRACT
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
A new norbornenyl-functionalized castor oil alcohol (NCA) was synthesized and ring-opening
metathesis copolymerized separately with two norbornene-based cross-linking agents:
dicyclopentadiene (DCPD) and a bifunctional norbornene crosslinker (CL). Isothermal differential
scanning calorimetry (DSC) was used to examine the cure behavior of NCA/DCPD and NCA/CL
resins, through which a reasonable cure schedule was determined. The glass transition temperature
(Tg) and storage modulus (E’), characterized by dynamic mechanical analysis (DMA), increased
significantly in both copolymer systems with the addition of cross-linking agents. Cross-link
density of the two systems was evaluated using a modified empirical equation from the kinetic
theory of rubber elasticity. Differences in tensile stress-strain behavior and thermal stability
between polymerized NCA/DCPD and NCA/CL were correlated to the structural rigidity and
cross-linking density resulting from the cross-linking agents.
Keywords:
Castor oil, Norbornenyl functionality, Cross-linking.
1

ACCEPTED MANUSCRIPT
1
Introduction
2
3
In recent decades, increasing concerns for the environment and sustainability have prompted
the exploration of bio-based, renewable plastics and composites for structural applications. Among
the candidates for renewable resources, agricultural oils have emerged as a powerful feedstock to
develop polymeric materials capable of replacing commercial polymer products currently derived
from petrochemcials. The polymerization and processing of agricultural oils into stiff, robust
materials in key applications, such as automotive components, civil engineering and construction
are on-going topics in both academic and industrial settings [1].
4
5
6
7
8
9
Hardening of drying oils, such as tung oil, readily occurs due to their inherently high extent
of unsaturation, but the majority of fatty acid chains of agricultural oils that possess few double
bonds that are not readily polymerizable (i.e., non-conjugated). Chemical modification of double
bonds and ester groups on triglycerides therefore is a powerful strategy to guide monomer design
towards controllable polymerization of agricultural oils. Through conversion of olefins to more
reactive functional groups, such as epoxy, hydroxyl or cyclic groups, the triglyceride is susceptible
to alternate polymerization paradigms or to subsequent functionalization [2]. A remarkable range
of polymers have been prepared from agricultural oils utilizing various mechanisms, including
free-radical polymerization, cationic polymerization, olefin metathesis polymerization, and
condensation polymerization, have been reviewed [3, 4].
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Many recent efforts have concentrated on optimization of the thermal and mechanical
properties of agricultural oil-based polymers. Because of the linear, flexible hydrocarbon chains
common in triglycerides, which are associated with high degree of chain rotation and multiple
segment movements, the resulting polymers usually exhibit low Tg and modulus [5]. Two
approaches to enhance the thermal and mechanical performance of these polymers exist. First,
polymer stiffness can be improved through introduction of “rigid” chemical moieties to the
polymer structure, such as aromatic, carbocyclic and heterocyclic groups [6]. Second, material
properties can be tuned by varying the level of covalent cross-linking in the polymer network. To
achieve this, the number of functional groups on the triglyceride molecules are commonly altered.
For soybean oil-based polymers, the hydroxyl functionalities per fatty chain can be manipulated
by chemical reactions such as glycerolysis, amidation, methanolysis, and ester reduction, which
endow the resulting polymers more cross-linking sites and significantly improved Tg and modulus
[7-9]. The addition of bifunctional, cross-linking comonomers from petroleum derivatives is
another suitable approach, although it partly compromises the bio-based content of the materials.
Divinylbenzene and norbornene-based cross-linkers have gained increased attention in optimizing
thermo-mechanical properties [10-13] because they serve to both cross-link the polymer and
incorporate rigidity to stiffen the bio-based polymers.
Ring-opening metathesis polymerization (ROMP) has recently proven promising for
developing strong and tough thermosets derived from vegetable oils. Due to the unique
characteristic of olefin retention during ROMP, the resultant polymers have the potential to serve
as reactive sites in the polymer matrix for self-healing composite applications [14]. Commercial
Ruthenium-based catalysts are capable of initiating polymerization of modified agricultural
oil-derived monomers due to such catalyst’s high reactivity, functional group tolerance, and
air/moisture insensitivity. Early work in ROMP of oils was explored by copolymerizing DilulinTM
[15], a commercially available norbornenyl-modified linseed oil derivative (Cargill Inc, MN), with
2

ACCEPTED MANUSCRIPT
1
2
various crosslinkers [16, 17]. However, the low number of norbornenyl groups (≤1 per triglyceride)
in Dilulin inhibits network formation with its comonomer, resulting in a plasticization effect that
decreases mechanical properties. To overcome the disadvantage of incomplete network formation,
norbornenyl-modified fatty alcohols derived by reacting 5-norbornene-2,3-dicarboxylate
anhydride with Dilulin and castor oil (NMDA and NMCA) were synthesized [7], resulting in an
average of 1.3 and 1.9 pendent norbornene groups per fatty chain, thereby improving storage
modulus and Tg without the need for comonomers. Another novel copolymer system was prepared
from norbornenyl-functionalized castor oil (NCO) and castor oil alcohol (NCA), which were
derived by reacting 5-norbornene-2-carbonyl chloride with castor oil and castor oil alcohol
respectively [8]. NCA monomer has approximately 1.8 norbornenyl groups per fatty chain, which
led to relatively short gelation time and high cross-link density. The molecular structure of the
aforementioned ROMP-reactive bio-monomers based on agricultural oils are listed in Fig. 1(a-e).
3
4
5
6
7
8
9
10
11
12
13
O
O
O
O
O
O
Dilulin
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Fig. 1. ROMP-reactive bio-monomers from Dilulin and castor oil;
(a) cyclopentadiene-modified linseed oil (Dilulin); (b) norbornenyl-modified Dilulin alcohol
(NMDA); (c) norbornenyl-modified castor oil alcohol (NMCA); (d) norbornenyl-functionalized
castor oil (NCO); and (e) norbornenyl-functionalized castor oil alcohol (NCA).
In this work, we investigate the effect of cross-linking functionality of ROMP-reactive
comonomers on the thermo-mechanical properties and the network structure of the NCA
thermosetting system. Two cyclic olefin comonomers, dicyclopentadiene (DCPD) and norbonene
crosslinker (CL) were ring-opening metathesis copolymerized with NCA monomer in varied ratios.
The cure behavior of NCA-DCPD and NCA-CL is tied to the ROMP reactivity of multifunctional
cross-linking agents [11, 16, 18]. Isothermal conversion and conversion rate of these two systems
during cure were monitored using differential scanning calorimetry (DSC). The
thermo-mechanical properties of cured copolymers, polyNCA/DCPD and polyNCA/CL, were
characterized using dynamic mechanical analysis (DMA). Glass transition temperature, storage
modulus, loss modulus and tangent peak served as useful metrics to probe structural network
changes. Mechanical properties obtained from tensile tests and thermal stability measured by
thermogravimetric analysis (TGA) are also discussed.
3

ACCEPTED MANUSCRIPT
1
2
2. Experimental
2.1 Materials
3
4
Castor oil, dicyclopentadiene (DCPD), lithium aluminum hydride (95%) (LAH),
5-norbornene-2-carboxylic acid (98%, a mixture of endo- and exo-, predominantly endo), thionyl
chloride (99.5%) (SOCl2), triethylamine (TEA) and 2nd generation Grubbs’ catalyst
[1,3-bis-(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro(phenylmethylene)(tricyclohexylp
hosphine) ruthenium] were obtained from Sigma-Aldrich (Milwaukee, WI). Tetrahydrofuran
(HPLC grade) (THF), ethyl acetate, methylene chloride (ACS certified) (CH2Cl2), anhydrous
magnesium sulfate (MgSO4), sodium hydroxide (NaOH), sodium carbonate (Na2CO3), and
hydrochloric acid (HCl) were supplied by Fisher Scientific (Fair Lawn, NJ). CH2Cl2 was refluxed
over CaH2 under nitrogen flow and distilled immediately prior to use. All other reagents were used
as received without further purification.
5
6
7
8
9
10
11
12
13
14
15
16
Norbornenyl-based
cross-linker
(CL),
a
mixture
of
exo-endo
isomer,
1,4,4a,5,8,8a-hexahydro- 1,4,5,8-exo,-endo-dimethanonaphthalene and endo-endo isomer,
1,4,4a,5,8,8a-hexahydro- 1,4,5,8-endo,-endo-dimethanonaphthalene in a molar ratio of 87:13, was
synthesized according to a literature method [19].
17
2.2. Synthesis of bicyclo[2.2.1]-hept-2-ene-5-carbonyl chloride
18
19
20
21
22
23
24
25
26
27
28
The preparation of bicyclo[2.2.1]-hept-2-ene-5-carbonyl chloride was modified from a
reported approach [20]: 5-norbornene-2-carboxylic acid (110.53 g, 0.8 mol) was mixed with
thionyl chloride (190.35 g, 1.6 mol) in a 500 mL round-bottom flask, and then refluxed under dry
nitrogen gas at 70oC for 3 h. Excess SOCl2 was evaporated under reduced pressure. The crude
product was further purified by distillation to eliminate colored impurity, yielding a transparent
1
liquid (71 %). H NMR (400Hz, CDCl3) indicates that the product is an mixture of endo isomer
1
(74%) and exo isomer (26%). H NMR δ: 6.25-6.27 (m, 1H, =CH, endo), 6.20-6.22 (m, 0.35H,
=CH, exo), 6.11-6.14 (m, 0.35H, =CH, exo), 6.02-6.04 (m, 1H, =CH, endo), 3.43-3.47 (m, 2H,
endo), 3.29 (s, 0.35H, exo), 2.99 (s, 1.4H), 2.72-2.76 (m, 0.35H, exo), 2.00-2.05(m, 0.36H, exo),
1.92-1.98 (m, 1.1H, endo), 1.47-1.55 (m, 2.9H), 1.41-1.43 (d, 0.36H, exo), 1.32-1.34 (d, 1.1H,
endo).
29
2.3. Synthesis of norbornenyl-functionalized castor oil alcohol (NCA)
30
31
32
33
34
35
36
37
38
NCA monomer was synthesized by following the reported route [19]: LAH (12.27 g, 0.32
mol) was added to a 2000 mL, two-neck round bottom flask with 100 mL of THF. Castor oil (100
g, 0.11 mol) was dissolved in 600 mL of THF and added dropwise to the LAH slurry in an ice bath.
The reaction was maintained at 0 oC for 12 h under mechanical stirring. To remove the remaining
LAH, the reaction mixture was quenched by ice water, and then 1M HCl was added until it
changed to a clear solution. The product was extracted with ethyl acetate (600 mL), washed with
water (600 mL), and then the organic layer was dried over MgSO4. After solvent evaporation, the
castor oil alcohol was obtained as a clear, light yellow viscous liquid (85%).
Castor oil alcohol (50g, 0.18 mol) was dissolved in 250 mL of anhydrous CH2Cl2 and cooled
4

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
to 0 oC in an ice bath. A solution of bicyclo [2.2.1] hept-2-ene-5-carbonyl chloride (57.8 g, 0.37
mol) in 250 mL of anhydrous CH2Cl2 was added dropwise and then followed by the addition of
triethylamine (39.13 g, 0.39 mol). The reaction was allowed to warm to ambient temperature and
the solution was stirred for 24 h. Then the reaction product was poured into 1000 mL of 5 wt.%
Na2CO3 aqueous solution and stirred for 12 h. Product was extracted with CH2Cl2, dried over
MgSO4 and removal of solvent yielded NCA as a light brown oil (80 %).
7
2.4. Sample preparation
8
9
Blends of NCA and cross-linking agents (DCPD or CL) were prepared at different
o
comonomer loadings: 10, 20, 30, 50 wt.%. CL (m.p. 12 C, determined by DSC) was mixed
o
10
11
12
13
14
15
16
17
18
19
20
21
22
23
directly at room temperature, while DCPD (m.p. 33 C) was melted in an oven prior to mixing
with NCA. Homogenous solutions were obtained for all blend compositions.
2nd generation Grubbs’ catalyst was freeze-dried prior to dissolution in monomer [21]. In a
typical example, catalyst (100 mg) was dissolved in 2 mL of benzene in a small vial and then
flash-frozen in a liquid nitrogen bath. The frozen solution was placed in an ice bath under vacuum
for 5 h to sublime the benzene. Freeze-dried 2nd generation Grubbs’ catalyst (0.125 wt.%) was
added to a resin mixture cooled in an ice bath, which was vigorously stirred with a spatula for 10
seconds to dissolve the catalyst completely. Small amounts of resin mixture (~100 mg) was
withdrawn for the cure kinetics study. The remaining resin was injected into rubber molds (100
mm × 70 mm × 1 mm) that were sandwiched between two glass plates. Specimens were cured
according to the following schedule (cure schedule developed vide infra, Section 3.1): isothermal
o
o
cure at 60 C for 1.5 h, followed by post-cure at 135 C for 12 h. After curing the resulting
thermosets were removed from the oven and cooled down in the air. The ROMP of NCA/DCPD
and NCA/CL mixtures is illustrated in Scheme 1.
5

ACCEPTED MANUSCRIPT
1
2
Scheme 1. ROMP of NCA with DCPD and CL.
3
4
2.5. Characterization
5
6
Differential scanning calorimetry (DSC, Q2000, TA instruments) was used to monitor
isothermal curing. Resin (~ 10 mg) was loaded into the DSC cell at a standby temperature of -50
oC. After thermal equilibration for 1 min, the DSC cell was heated to the required temperature (Tc
7
o
o
8
= 50, 60, 70 C) at a heating rate of 100 C/min and held isothermally for 1 h. Resulting cured
o
9
samples were cooled to -50 C at a rate of 10oC/min, followed by a dynamic scan to 250 oC at a
10
11
12
13
14
15
16
17
rate of 10 oC/min to reach full conversion.
Soxhlet extraction was performed on ca. 2.5 g samples with 100 mL of refluxing CH2Cl2 in a
Soxhlet extractor for 24 h. After extraction both the remaining solid and the extracts were dried in
a vacuum oven at 60 oC overnight.
Swelling tests of polyNCA/DCPD and polyNCA/CL were performed on small rectangular
pieces (1 × 10 × 20 mm). Pre-weighed samples (Winitial) were soaked in CH2Cl2 for 48 h at room
temperature. The final weight (Wfinal) of saturated samples was used in equation 3 to calculate
swelling percentage:
Wfinal −Winitial
18
Swelling(%) =
×100%
(1)
Winitial
19
20
21
22
23
Dynamic mechanical analysis (DMA, Q800, TA Instruments) was performed on rectangular
sample strips of 20 × 4.3 × 0.8mm (length × width × thickness), measured in a tension fixture,
with uniaxial oscillation at a constant frequency of 1 Hz and amplitude of 15 ꢀm. The samples
were scanned at a temperature range of -50 oC to 200 oC at a heating rate of 3 oC /min.
The tensile properties of the thermosetting copolymers were measured using an Instron
6

ACCEPTED MANUSCRIPT
1
2
universal testing machine (model 4502) with a crosshead speed of 50 mm/min. Testing specimens
(ASTM D 638-03, type V) were machine-milled, and at least three replicates of each sample were
taken for measurement. Young’s modulus (E), maximum tensile strength (σm), strain at break (εb),
and tensile toughness of the polymers were recorded.
3
4
5
Thermal stability was investigated using a thermogravimetric analyzer (TGA) Q50 (TA
Instruments). PolyNCA/DCPD and polyNCA/CL samples (~10 mg) were heated from ambient
temperature to 700 oC at a heating rate of 20 oC /min in air (50 mL/min).
6
7
8
9
10
3. Results and discussion
11
3.1 Isothermal cure kinetics
o
12
13
14
15
16
17
Isothermal curing was performed at three temperatures 50, 60, and 70 C for NCA resins
blended with DCPD or CL at a 50:50 mass ratio. Figs. 2(a) and 2(b) show the exothermic peaks
corresponding to isothermal cure. These curing temperatures were selected on the basis of the
onset temperature of exothermic peak determined from standard dynamic scans of NCA resin
blends.
Conversion, α, or degree of cure, is defined as:
∆Ht
18
α
=
(2)
∆Hrxn
19
20
21
where ꢀHt is the heat generated up to time t during the cure reaction and ꢀHrxn is the total reaction
heat per unit mass (J/g) from uncured (α = 0) to fully cured (α = 1). The conversion rate is defined
as:
d
α
dH / dt
∆Hrxn
22
=
(3)
dt
23
24
25
26
27
28
29
30
31
32
33
34
35
36
where dH/dt represents the heat flow rate by DSC.
The NCA50-DCPD50 and NCA50-CL50 blends show the similar times for the maximum
conversion rate to be reached. For both resin compositions, the time taken for maximum
conversion rate reduced significantly from about 2.5 min to about 12 s when the cure temperature
o
o
was raised from 50 C to 70 C, exhibiting a significant influence of cure temperature on the
ROMP cure kinetics. Compared to NCA/DCPD, the maximum conversion rate of NCA/CL resin
is higher under the same curing condition, confirming a higher ROMP reactivity of CL than
DCPD [22].
Figs. 3(a) and 3(b) show conversion with respect to time for both resins at different
temperatures. The final conversion obtained after one-hour isothermal cure is shown in Table 1. At
70 oC, 90 % of NCA/DCPD monomers were converted into the polymeric network in only 5 min,
while NCA/CL reacted even faster. However, the fact that the conversion levels off below
complete conversion means that a post-cure process, where the temperature is set above Tg∞ (glass
transition temperature in the fully cured state) is necessary to fully postcure the polymer network.
7

ACCEPTED MANUSCRIPT
1
(a)
(b)
2
3
4
5
Fig. 2. Isothermal DSC thermogram of (a) NCA50/DCPD50 and (b) NCA50/CL50 at different
curing temperatures.
8

ACCEPTED MANUSCRIPT
(a)
1
(b)
2
3
4
5
6
Fig. 3. Conversion as a function of time of (a) NCA50/DCPD50 and (b) NCA50/CL50 under
isothermal cure at different curing temperatures.
9

ACCEPTED MANUSCRIPT
1
2
Table 1.
Selected isothermal cure data from Fig. 2 and Fig. 3.
Composition
NCA50/DCPD50
NCA50/CL50
Curing
Time of (dH/dt)
max (s)
(dH/dt) max
(W/g)
Final conversion
α (%)
temperature (oC)
50
60
70
50
60
70
168 ± 2
0.53 ± 0.02
6.1 ± 0.20
12 ± 0.5
84.4 ± 1.4
93.6 ± 1.2
95.5 ± 1.2
87.8 ± 1.1
90.5 ± 1.5
97.7 ± 1.2
42.0 ± 0.6
12.3 ± 0.3
149 ± 1
1.5 ± 0.05
7.1 ± 0.30
95 ± 3.1
46.5 ± 0.3
12.5 ± 0.4
3
4
5
6
7
8
9
Consequently, a cure schedule (described in Section 2.4) was proposed to fabricate bulk
specimens for NCA resins with various levels of DCPD or CL loading. The curing temperature
o
was set as 60 C for all compositions as a trade-off cure condition for resins with higher
comonomer loading (> 30 %) while ensuring high conversion rates for low comonomer loading(<
30 %). A post-cure temperature higher than Tg∞ (135 oC) was selected to ensure full curing of resin
without thermal degradation.
10
3.2 Glass transition temperature
11
12
13
14
15
16
17
Fig. 4 shows the evolution of glass transition temperature (Tg) as a function of cross-linking
agent loading for the NCA/comonomer thermosets. The value of Tg here is determined by the
maximum of dissipation factor peak from DMA measurements. It has been demonstrated that DSC
is not sensitive to detect Tg accurately for NCA-dominant copolymers (NCA > 80 %), likely due to
the effect of cross-linking and the multiple relaxation modes of fatty acid-based polymer chains
[23].
18
19
20
21
22
23
Fig. 4. Glass transition temperature (Tg) of polyNCA/DCPD and polyNCA/CL as a function of
cross-linking agent loading.
In Fig. 4, polyNCA/DCPD and polyNCA/CL curves display a monotonic increase of Tg with
the addition of cross-linking agents, with the value ranging from 52.3 oC for polyNCA to 101.8 oC
10

ACCEPTED MANUSCRIPT
o
1
2
and 134.3 C for polyNCA/DCPD and polyNCA/CL, respectively, which is consistent with
previous reports of copolymerization of linseed oil with DCPD or CL [11, 16, 24]. In this study,
the effect of DCPD and CL on the Tg shift behavior of the castor oil alcohol-based polymer is
evaluated regarding the aforementioned two aspects: structural rigidity contribution and the
crosslink density.
3
4
5
6
The Fox equation is a simple empirical model to predict the dependence of Tg on composition
in a binary polymer system, usually polymer blends and copolymers. The Tg of neat polyNCA was
measured as 52.3 oC and polyDCPD prepared by ROMP was reported as 153.9 oC previously [25].
Since the Tg of polyCL is not available yet for the difficulty of preparation, here only the Tg of
polyNCA/DCPD is calculated as a function of its mass ratio based on the Fox equation,
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
1
w
w2
1
=
+
(4)
Tg Tg1 Tg 2
Tg1 and Tg2 represent the glass transition temperatures of polyNCA and polyDCPD, respectively,
while w1 and w2 are the corresponding mass fractions. A positive deviation of experimental Tg
from Fox’s prediction beyond error evaluation was observed for polyNCA/DCPD. The Tg
increasing trend with DCPD loading differs from the linear relationship that was reported in
previous NCO/NCA bio-polymer systems [8]. Nonetheless, it presented similarity to the Tg
behavior of ROMP thermosets based on norbornene petroleum-derivatives with high cross-link
density [26]. This observed deviation of Tg from the model could therefore be related to the high
cross-linking density resulting from the high functionality of the NCA-comonomer compositions.
Unlike Dilulin or NCO-based copolymers, of which the norbornenyl functionality per molecule is
about 0.8 ~ 1, the NCA monomer affords a high cross-linked network by itself due to 1.8
ROMP-reactive norbornene rings per molecule (f = 1.8). As the cross-linking comonomer, CL has
two norbornene rings of equal reactivity (f = 2), while DCPD (in terms of endo- isomer) is able to
lightly cross-link due to 20 % of cyclopentene ring opening except for norbornene ring opening (f
≈ 1.2) [27]. It is expected that a low-molecular-weight comonomer with high functionality, such as
DCPD and CL, would tune the cross-link density of NCA around a high level. The altered Tg -
composition dependence of polyNCA/DCPD from the model may reflect the reduction of free
mobility of chain segments by the strong interaction in such a highly cross-linked system.
Besides, the higher Tg shown in polyNCA/CL compared to polyNCA/DCPD at the same
comonomer loading is also associated with CL’s higher functionality. CL retains a bicyclic
structure after both norbornene groups undergoes ROMP and the structural units are theoretically
all incorporated into copolymer backbone (in Scheme 1 (ii)). DCPD, on the other hand, is
polymerized leaving with about 80 % pendant cyclopentene rings, which gives lower contribution
of rigid cyclic structure to the main polymer chains (in Scheme 1 (i)). As a result, CL appears to
be a more efficient cross-linking agent than DCPD to improve Tg, as supported by the increasing
difference of Tg between polyNCA/CL and polyNCA/DCPD with comonomer loading.
39
3.3 Thermo-mechanical properties
40
41
42
Fig. 5(a) and 5(b) shows the storage modulus and tan δ as a function of temperature for
polyNCA/DCPD and polyNCA/CL samples. The E' of polyNCA/DCPD at 50 % DCPD loading is
1514 ± 180 MPa at room temperature, an increase of 2.8 times that of neat polyNCA.
11

ACCEPTED MANUSCRIPT
1
2
Comparatively, E' of polyNCA/CL at 50 % CL loading increased 2 times that of neat polyNCA,
reaching up to 2046 ± 46 MPa. The significant modulus improvement in the glassy state results
from the shift of the glass-rubber transition to higher temperatures. The nature of the sharp E'
decrease, indicative of the glass transition, relative to cross-linking agent loading, differs
considerably for polyNCA/DCPD and polyNCA/CL between. The E' decrease for
polyNCA/DCPD appears to occur over a narrower temperature window and with a larger
magnitude with increasing comonomer loading than that of polyNCA/CL. Interestingly for
polyNCA/CL, further raising the CL loading from 30 % towards 50 %, the slope of the E' decrease
reverts to the shape resembling the neat polyNCA.
3
4
5
6
7
8
9
10
(a)
11
(b)
12
13
Fig. 5. Storage modulus and tan
δ curves for (a) polyNCA/DCPD and (b) polyNCA/CL with
14
15
16
17
18
increasing cross-linking agent loading.
Corresponding to the shift of E' decrease, the tan
δ
curve of polyNCA/DCPD in Fig 5 (a)
shows a single peak that becomes sharper and taller with increasing DCPD loading. A
o
o
low-temperature and broad (between -20 C and 40 C) dissipation shoulder is present for both
12

ACCEPTED MANUSCRIPT
1
2
neat polyNCA and copolymers thereof, which decreases in intensity with the addition of DCPD.
This shoulder indicates a broad, sub-glass transition in polyNCA, which was previously attributed
to local crankshaft motions of ester groups between fatty chains and the conformational changes
of the cyclopentane rings [28]. The diminishing intensity of sub-glass relaxation of
polyNCA/DCPD results directly from the decrease in the NCA fraction. In Fig 6(a), the E” of
polyNCA/DCPD displays a similar decreasing dissipation shoulder with increasing DCPD loading.
Meanwhile, the shoulder separates from the primary transition peak as Tg shifts to higher
temperature.
3
4
5
6
7
8
9
For polyNCA/CL, the tan
δ
curve shown in Fig. 5(b) develops with CL loading in a
peak appears to span in broader
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
complicated manner. Compared to polyNCA/DCPD, the tan
δ
temperature range and shrink in peak intensity by increasing CL loading up to 30 %, which
indicates the formation of a complex thermosetting network structure with high inhomogeneity.
Three identifiable relaxation regions are observed in the tan
δ peak with the highest 50 % CL
loading and assigned to low-, medium- and high-temperature region for convenience. The E''
curves plotted on a linear scale in Fig. 6(b) demonstrates these regions (pointed by black arrows)
with better visibility. The E'' peak of the low-temperature relaxation region is stably located
o
around 20 C without changing with CL loading. With comparable results to the dissipation
shoulder of polyNCA/DCPD, the low-temperature relaxation region of polyNCA/CL is also
attributed to the sub-glass transition owing to NCA component. The medium- and
high-temperature regions are expected be associated with the primary relaxations of multiple
phases in the copolymer network, because the tan
temperature with the increase of CL loading. Last but not least, the decrease in intensity of the
primary tan peak for polyNCA/CL discontinues at 30 % CL loading while the peak remains
shifting to higher temperatures. Further increase of CL loading up to 50 % alters the evolving
pattern of tan curves, which is consistent with the change in the glass-transitional E' decrease. In
order to better explain the dynamic mechanical behavior of polyNCA/CL and polyNCA/DCPD, a
δ (or E'') peaks of both regions shift to higher
δ
δ
correlation between cross-link density, Tg, and tan δ peak is established.
(a)
29
13

ACCEPTED MANUSCRIPT
(b)
1
2
3
4
5
Fig. 6. Loss modulus curves of (a) polyNCA/DCPD and (b) polyNCA/CL with increasing
cross-linking agent loading.
6
3.4 Evaluation of Cross-Link Density
7
8
Fig. 5(a) and 5(b) show that in the rubbery plateau region, at temperatures above the glass
transition, copolymers begin to show elastic characteristics. The crosslink density can be generally
determined from the equilibrium shear modulus, which is commonly obtained at Tg + 50 oC in the
rubbery plateau. The kinetic theory of rubber elasticity provides the following relationship [29]:
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
ΦdRT
d
G =
ν e =
(5)
Mc
Mc
where Φ is the front factor, d is the density of the polymer, R is the gas constant, and T is the
o
absolute temperature at Tg + 50 C. Mc is the average molecular weight of elastically active
network chains (EANCs) between cross-links, which is inversely proportional to the crosslink
density
νe. The equation is only valid for lightly cross-linked rubbers that comply with the
assumptions of Gaussian chains. Φ is assumed to be unity only when G < 10 MPa [30]. For highly
cross-linked polymers, as is the case in this work, the EANC between two junctions are much
shorter and less likely to be randomly curled. It was reported that Equation 5 does not give
accurate predictions for highly cross-linked polymers such as epoxy and phenol-formaldehyde
resins [31]. Hence, it has been suggested that when elastic shear modulus values exceed 10 MPa,
showing a characteristic of non-Gaussian polymer chains, a modified empirical equation should be
used to calculate cross-link density and is given as:
(6)
27
logG = 6+ 23.6dMx / Mc
28
29
30
where Mx is the molecular weight of the multifunctional cross-linked atoms plus the attached
hydrogen atoms. The shear modulus can be substituted by the storage modulus, because for ideal
14

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
rubber: G = E'/3. The density of polymers is unified for the sake of calculation. Cross-link density
νe and molecular weight between cross-links Mc of polyNCA/DCPD and polyNCA/CL are listed
in Table 2. The calculated results are consistent with reasonable expectations: for example, Mc of
neat polyNCA agrees with the molecular weight of NCA monomer unit (C34H52O4, MW: 524.8).
Table 2.
Dynamic mechanical properties and cross-link density determined by DMA.
Tg (oC)
Cross-link Density
Soluble
Fraction
(%)
Copolymer
E' at 30 oC
tan δ peak
E'' peak
Mc
tan
δ
peak
νe
Composition
max
(MPa)
(mol/m3)
(g/mol)
NCA100
13.4 ± 1.2
21.9 ± 0.7
58.6 ± 0.3
73.8 ± 1.1
92.9 ± 0.7
56.3 ± 7.9
56.4 ± 3.6
60.5 ± 1.8
113.0 ± 1.7
52.3 ± 6.6
60.7 ± 0.2
72.6 ± 0.1
85.2 ± 0.3
101.8 ± 3.2
79.4 ± 4.3
98.1 ± 2.6
108.6 ± 0.4
134.3 ± 0.3
0.31 ± 0.01
0.34 ± 0.03
0.32 ± 0.02
0.46 ± 0.01
0.53 ± 0.02
0.18 ± 0.01
0.15 ± 0.01
0.14 ± 0.01
0.35 ± 0.01
537 ± 86
2014
1935
1875
1573
1454
2554
2689
2627
1759
497
1.1
NCA90/DCPD10
NCA80/DCPD20
NCA70/DCPD30
NCA50/DCPD50
NCA90/CL10
860 ± 45
517
533
636
689
392
372
381
569
-
1120 ± 57
1312 ± 75
1455 ± 78
1109 ± 58
1340 ± 60
1473 ± 33
2046 ± 46
-
-
0.9
-
NCA80/CL20
-
NCA70/CL30
-
NCA50/CL50
1.0
8
9
As can be seen, the cross-link density decreases monotonically with the addition of DCPD,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
and the value of Mc increases from 497 g/mol to 689 g/mol. While this general trend is in
agreement with the observed magnitude change in the relaxation peak in tan δ, it is noteworthy
that Tg exhibits a non-linear increase with DCPD loading. The inverse relationship between Tg and
cross-link density in polyNCA/DCPD system does not abide by the normal results from previous
bio-thermoset studies. The decrease of cross-link density, due to the modification of
multi-functionality between NCA and DCPD, is expected to have negative effect on Tg for the
resulting thermosetting system. In fact, the Tg increase of polyNCA/DCPD appears to be
dominantly influenced by increased structural rigidity rather than cross-linking effects. The main
contributors, cyclopentane group and carbon-carbon double bond, which are characteristic in
ROMP polymers based on norbornene-derivatives, endow chain segments with very low
compliance and high activation energy for macroscopic cooperative movement. With increasing
DCPD loadings, the thermoset network contained increased cyclic rigidity in lieu of the flexible
backbone of the NCA fraction, resulting in an increase in Tg.
While investigating polyNCA/CL at a relatively low CL loading (< 30 wt.%), a “typical”
cross-linking effect for thermosets is observed. As the cross-link density increases, the tan
δ peak
representing the glass-to-rubber relaxation broadens, diminishes in intensity and shifts to higher
temperature. The corresponding drop in E' also decreases in magnitude and expands over a wider
temperature range. The increasing rate of
finally reaches a plateau between 20 wt.% and 30 wt.%. When the CL loading is further increased
to 50 wt.%, e appears to decrease back to a value lower than polyNCA. The swelling percentage
shown in Fig. 7 gives a qualitative interpretation of e which correlates well to the calculated e of
νe, however, slows down with the addition of CL and
ν
ν
ν
polyNCA/CL. The impaired cross-linking effect by high CL loading is also manifested in the
15

ACCEPTED MANUSCRIPT
1
2
intensified E' decrease and tan δ relaxation peak indicating the transition of polyNCA/CL.
Counterintuitive cross-linking trends have been observed previously with CL-copolymerized
ROMP systems. In a Dilulin-CL copolymer, νe leveled off at approximately 40 wt.% of CL
loading [11]. In the case of a DCPD/CL copolymer, the effect of CL loading on νe became
negligible between 15 and 20 wt.% [32]. These phenomena could be related to reaction-induced
microphase separation as a consequence of demixing on incipient formation of ROMP copolymer
with high CL content [32]. At the initial curing stage, highly reactive CL is likely to form
individual cross-linked microgels before the formation of a cohesive network. After the reaction
rate of CL significantly decreases as CL depletes, the propagation of NCA dominates the network
growth till fully cured. This interpretation is evidenced by the multiple relaxations shown in the
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
tan
δ
peak (Fig. 5(b)) and E” peaks (Fig. 6(b)) of cured polyNCA/CL. In addition to the sub-Tg
o
transition at around 10 ~ 40 C, there are two primary transitions responsible for the multiple
phases: the high Tg phase and the low Tg phase, according to the previous identification of the
relaxation regions. The Tg of high Tg phase, determined by tan
increases substantially with CL loading. Hence, the high Tg phase is likely identified as CL-rich
microgels. The Tg of low Tg phase is indicated by tan peak in the medium-temperature region
and slowly shifts from 60 oC to 80 oC, which ascribes the low Tg phase to NCA-rich fraction. The
generation of network heterogeneity, likely due to the differences in ROMP reactivity between
NCA and CL, reduces the macroscopic cross-link density of the thermosets [33].
δ max in the high-temperature region,
δ
The soluble oligomer fraction determined by Soxhlet extraction was present in a few
copolymer formulations, very low (around 1 wt.%, in Table 2), and independent of νe for both
polyNCA/DCPD and polyNCA/CL. This implies that cross-linking between NCA and the
cross-linking agents is efficient and most monomer is incorporated into the polymer network.
25
26
27
28
Fig. 7. Weight-based swelling percentage of polyNCA/DCPD and polyNCA/CL as a function of
cross-linking agent loading.
29
3.5 Tensile Properties
30
31
Fig. 8 illustrates the stress-strain behavior of polyNCA/DCPD and polyNCA/CL, determined
by tensile testing. Two mass ratios (30/70 and 50/50) of each copolymer were compared to the
16

ACCEPTED MANUSCRIPT
1
2
neat polyNCA by investigating Young’s modulus, maximum tensile strength, strain at break, and
toughness (in Table 3). The two copolymers exhibited a different variation in tensile behavior with
the addition of the cross-linking agents. Neat polyNCA exhibits ductile behavior with a yield point
that occurs at about 2 % strain, followed by plastic deformation and breaking at 12.4 ± 1.8 %
strain. As DCPD loading increases to 30 % and 50 %, a considerable improvement in Young’s
modulus, tensile strength, and toughness is observed for polyNCA/DCPD copolymers.
PolyNCA50/DCPD50 reaches the highest tensile properties: Young’s modulus of 1699 ± 428 GPa,
tensile strength of 42.9 ± 0.6 MPa, toughness of 4.95 ± 0.55 J/m3, and elongation at break (14.0 ±
1.6 %). The improved tensile properties can be attributed to the effect of increasing Tg as well as
declining cross-linking density by incorporating DCPD into polyNCA. Incorporation of a larger
percentage of rigid cyclopentane rings into the backbone chains strengthens covalent and
intermolecular bonding as well as the polymer’s resistance upon stress, while the reduced
cross-linking density is responsible for the larger elongation at break.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
In comparison, polyNCA/CL copolymers exhibit lower strain at break in addition to increases
in tensile strength and modulus when increasing CL loadings, demonstrating a trend to brittle
failure. PolyNCA50/CL50 displays a slight increase in εb (by 3.6 ± 0.5) and toughness (by 2.23 ±
0.30 J/m3) from polyNCA70/CL30, while E and σt continue to increase with CL loading. The shift
of stress-strain curves according to CL loading is also influenced by the two variables of Tg and νe
but in a different way compared to polyNCA/DCPD system. The addition of CL increasing both
Tg and cross-link density performs more efficient improvement on the tensile strength and
modulus of polyNCA/CL, however detrimental to plastic deformation.
23
24
25
26
27
28
29
Fig. 9. Tensile stress-strain curves of polyNCA/DCPD and polyNCA/CL at various cross-linking
agent loadings.
Table 3.
Tensile properties of polyNCA/DCPD and polyNCA/CL.
17

ACCEPTED MANUSCRIPT
Composition
NCA100
E (MPa)
σm (MPa)
15.5 ± 1.4
28.2 ± 0.5
42.9 ± 0.6
38.9 ± 0.2
52.2 ± 0.2
εb (%)
Toughness (J/m3)
1.37 ± 0.32
2.66 ± 0.43
4.95 ± 0.55
1.36 ± 0.10
3.59 ± 0.40
469 ± 47
1213 ± 178
1699 ± 428
1639 ± 259
2051 ± 128
12.4 ± 1.8
11.6 ± 1.9
14.0 ± 1.6
5.2 ± 0.3
8.8 ± 0.8
NCA70DCPD30
NCA50DCPD50
NCA70CL30
NCA50CL50
1
2
E = Young’s modulus, σm = maximum tensile strength, εb = elongation at break
3.6 Thermal stabilities
3
4
Thermogravimetric analysis results for polyNCA/DCPD and polyNCA/CL are shown in Fig.
9(a) and 9(b). Assisted by the weight loss derivative, all TGA curves for the two groups of
copolymers can be divided into three stages. The first stage represents a slow decline over the
5
o
6
temperature range from 250 to 400 C, which is probably related to the decomposition of side
7
chains or branches that are lightly cross-linked within the thermosetting network. Soluble fractions
or unreacted monomers may not be an influential factor to this stage because the Soxhlet
extraction results evidenced that the resins were fully cured. The second stage ranging from 400 to
8
9
o
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
500 C corresponds to a significant collapse in weight percentage. During this stage, the
thermosets are subjected to fast decomposition of main chains in the network, the cleavage of C=C
bonds and ester groups for instance. The third stage, which extends to 650 oC, is responsible for
the oxidation of residual char from the last stage.
As shown in Fig. 9(a), the weight loss curve is elevated stepwise for all three stages with the
increase of DCPD loading in polyNCA/DCPD. The cross-linking of rigid cyclic moiety from
DCPD imparts better thermal stability to the bio-based NCA polymer. Fig. 9(b) shows that
polyNCA/CL does not exhibit such gradient shift with CL loading to polyNCA/DCPD. The weight
loss of polyNCA/CL appear below neat polyNCA during the first degradation stage, as illustrated
in the inset graph. The reason can be due to the formation of pronounced heterogeneity in the
polyNCA/CL network. In a less homogeneous polyNCA/CL compared to polyNCA/DCPD, higher
number of dangling chains that are loosely tied to the main network may be subjected to
de-bonding and oxidization at early elevated temperature. PolyNCA/CL’s resistance to
degradation is therefore limited due to its network heterogeneity.
(a)
25
18

ACCEPTED MANUSCRIPT
(b)
1
2
3
Fig. 9. Thermogravimetric analysis of (a) polyNCA/DCPD and (b) polyNCA/CL with increasing
cross-linking agent loading.
4
4. Conclusions
5
6
Norbornenyl-functionalized castor oil alcohol (NCA) was prepared and copolymerized with
two cross-linking agents: DCPD and CL. CL exhibits higher ROMP reactivity than DCPD in
curing with NCA. With increasing loading of cross-linking agents, the glass transition temperature,
storage modulus and tensile properties of both polyNCA/DCPD and polyNCA/CL increases
significantly in a different manner, which is ascribed to the molecular structure and functionality
of the cross-linking agents. The glass transition region of the thermosets, however, exhibits a
complex dependence on the cross-linking agent loading. This is attributed to changes in cross-link
density concomitant with the enhanced structural rigidity. These two factors also influence the
ductile/brittle tensile behaviors of polyNCA/DCPD and polyNCA/CL. Structural heterogeneity
causes the diminished thermal stability of polyNCA/CL compared to polyNCA/DCPD.
7
8
9
10
11
12
13
14
15
Acknowledgment
16
17
18
19
20
21
22
This material is based upon work supported by the National Science Foundation under Grant No.
CMMI1348747. We are also thankful to Professor Richard Larock from the Department of
Chemistry at Iowa State University (ISU) for the use of lab equipment for monomer preparation.
Additionally, we thank Dr. David Grewell and Mr. Russell Huffman from the Department of
Agricultural and Biosystems Engineering at ISU for their careful assistance in the preparation of
tensile testing samples.
19

ACCEPTED MANUSCRIPT
REFERENCES
1.
2.
Lligadas G, Ronda JC, Galia M, and Cadiz V. Materials Today 2013;16(9):337-343.
Khot SN, Lascala JJ, Can E, Morye SS, Williams GI, Palmese GR, Kusefoglu SH, and Wool RP.
Journal of Applied Polymer Science 2001;82(3):703-723.
3.
4.
5.
Xia Y and Larock RC. Green Chemistry 2010;12(11):1893.
Meier MAR, Metzger JO, and Schubert US. Chemical Society Reviews 2007;36(11):1788-1802.
Wool RP. 4 - Polymers and Composite Resins from Plant Oils. In: Wool RP and Sun XS, editors.
Bio-Based Polymers and Composites. Burlington: Academic Press, 2005. pp. 56-113.
Can E, Kusefoglu S, and Wool RP. Journal of Applied Polymer Science 2002;83(5):972-980.
Xia Y, Lu YS, and Larock RC. Polymer 2010;51(1):53-61.
6.
7.
8.
9.
Xia Y and Larock RC. Polymer 2010;51(12):2508-2514.
Jeong W, Mauldin TC, Larock RC, and Kessler MR. Macromolecular Materials and Engineering
2009;294(11):756-761.
10.
11.
Badrinarayanan P, Lu YS, Larock RC, and Kessler MR. Journal of Applied Polymer Science
2009;113(2):1042-1049.
Mauldin TC, Haman K, Sheng X, Henna P, Larock RC, and Kessler MR. Journal of Polymer
Science Part a-Polymer Chemistry 2008;46(20):6851-6860.
12.
13.
14.
Kundu PP and Larock RC. Biomacromolecules 2005;6(2):797-806.
Li FK, Hasjim J, and Larock RC. Journal of Applied Polymer Science 2003;90(7):1830-1838.
Hondred PR, Salat L, Mangler J, and Kessler MR. Journal of Applied Polymer Science
2014;131(12).
15.
16.
17.
Chen JX, Soucek MD, Simonsick WJ, and Celikay RW. Polymer 2002;43(20):5379-5389.
Henna P and Larock RC. Journal of Applied Polymer Science 2009;112(3):1788-1797.
Mauldin TC, Haman K, Sheng X, Henna P, Larock RC, and Kessler MR. Journal of Polymer
Science Part A: Polymer Chemistry 2008;46(20):6851-6860.
18.
Henna PH, Kessler MR, and Larock RC. Macromolecular Materials and Engineering
2008;293(12):979-990.
19.
20.
Stille JK and Frey DA. Journal of the American Chemical Society 1959;81(16):4273-4275.
Biswas S, Belfield KD, Das RK, Ghosh S, and Hebard AF. Chemistry of Materials
2009;21(23):5644-5653.
21.
Jones AS, Rule JD, Moore JS, White SR, and Sottos NR. Chemistry of Materials
2006;18(5):1312-1317.
22.
23.
24.
Sheng X, Lee JK, and Kessler MR. Polymer 2009;50(5):1264-1269.
Can E, Wool RP, and Küsefoğlu S. Journal of Applied Polymer Science 2006;102(3):2433-2447.
Xia Y, Henna PH, and Larock RC. Macromolecular Materials and Engineering
2009;294(9):590-598.
25.
26.
Jeong W and Kessler MR. Chemistry of Materials 2008;20(22):7060-7068.
Sheng X, Kessler MR, and Lee JK. Journal of Thermal Analysis and Calorimetry
2007;89(2):459-464.
27.
28.
29.
Matejka L, Houtman C, and Macosko CW. Journal of Applied Polymer Science
1985;30(7):2787-2803.
Madbouly SA, Xia Y, and Kessler MR. Macromolecular Chemistry and Physics
2013;214(24):2891-2902.
Levita G, Depetris S, Marchetti A, and Lazzeri A. Journal of Materials Science
20

ACCEPTED MANUSCRIPT
1991;26(9):2348-2352.
30.
31.
Tobolsky AV, Takahashi M, Schaffha.R, and Katz D. Journal of Polymer Science Part a-General
Papers 1964;2(6pa):2749-2758.
Nielsen LE. Journal of Macromolecular Science-Reviews in Macromolecular Chemistry 1969;C
3(1):69-103.
32.
33.
Yang YS and Lee LJ. Polymer 1988;29(10):1793-1800.
Sahagun CM and Morgan SE. Acs Applied Materials & Interfaces 2012;4(2):564-572.
Figures and Tables
Scheme 1. ROMP of NCA with DCPD and CL.
Fig. 1. ROMP-reactive bio-monomers from Dilulin and castor oil; (a) cyclopentadiene-modified
linseed oil (Dilulin); (b) norbornenyl-modified Dilulin alcohol (NMDA); (c)
norbornenyl-modified castor oil alcohol (NMCA); (d) norbornenyl-functionalized castor oil
(NCO); and (e) norbornenyl-functionalized castor oil alcohol (NCA).
Fig. 2. Isothermal DSC thermogram of (a) NCA50/DCPD50 and (b) NCA50/CL50 at different
curing temperatures.
Fig. 3. Conversion as a function of time of (a) NCA50/DCPD50 and (b) NCA50/CL50 under
isothermal cure at different curing temperatures.
Fig. 4. Glass transition temperature (Tg) of polyNCA/DCPD and polyNCA/CL as a function of
cross-linking agent loading.
Fig. 5. Storage modulus and tan
δ curves for (a) polyNCA/DCPD and (b) polyNCA/CL with
increasing cross-linking agent loading.
Fig. 6. Loss modulus curves of (a) polyNCA/DCPD and (b) polyNCA/CL with increasing
cross-linking agent loading.
Fig. 7. Weight-based swelling percentage of polyNCA/DCPD and polyNCA/CL as a function of
cross-linking agent loading.
Fig. 8.
Tensile stress-strain curves of polyNCA/DCPD and polyNCA/CL at various
cross-linking agent loadings.
Fig. 9. Thermogravimetric analysis of (a) polyNCA/DCPD and (b) polyNCA/CL with
increasing cross-linking agent loading.
Table 1 Selected isothermal cure data from Fig. 3 and Fig. 4.
Table 2 Dynamic mechanical properties and cross-link density determined by DMA.
Table 3 Tensile properties of polyNCA/DCPD and polyNCA/CL.
21