Novel long chain unsaturated diisocyanate from fatty acid: Synthesis, characterization, and application in bio-based polyurethane

Article abstract of DOI:10.1002/pola.24114

A novel long chain linear unsaturated terminal diisocyanate, 1,16-diisocyanatohexadec-8-ene (HDEDI) was synthesized from oleic acid via Curtius rearrangement. Its chemical structure was identified by FTIR, ¹H NMR, ¹³C NMR, and HRMS. This diisocyanate was used as a starting material for the preparation of entirely bio-based polyurethanes (PUs) by reacting it with canola diol and canola polyol, respectively. The physical properties and crystalline structure of the PUs prepared from this diisocyanate were compared to their counterparts prepared from similar fatty acid-derived diisocyanate, 1,7-heptamethylene diisocyanate (HPMDI). The HDEDI based PUs demonstrated various different properties compared to those of HPMDI based PUs. For example, HDEDI based PUs exhibited a triclinic crystal form; whereas HPMDI based PUs exhibited a hexagonal crystal lattice. In addition, canola polyol-HDEDI PU demonstrated a higher tensile strength at break than that of canola polyolHPMDI, attributed to the higher degree of hydrogen bonding associated with the former sample. Nevertheless, lower Young's modulus and higher elongation in canola polyol-HDEDI PU were obtained because of the flexibility of the long chain introduced by the HDEDI diisocyanate.

Full text of DOI:10.1002/pola.24114

Novel Long Chain Unsaturated Diisocyanate from Fatty Acid: Synthesis,
Characterization, and Application in Bio-Based Polyurethane
1
2
1
LEILA HOJABRI, XIAOHUA KONG, SURESH S. NARINE
1
Department of Physics, Astronomy, and Chemistry, Trent Biomaterials Research Program, Trent University, Peterborough,
Ontario, Canada K9J 7B8
2
Department of Agricultural, Food and Nutritional Science, University of Alberta, 4-10, Agriculture/Forestry Centre, Edmonton,
Alberta, Canada T6G 2P5
Received 19 November 2009; accepted 29 April 2010
DOI: 10.1002/pola.24114
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: A novel long chain linear unsaturated terminal diiso-
cyanate, 1,16-diisocyanatohexadec-8-ene (HDEDI) was synthe-
sized from oleic acid via Curtius rearrangement. Its chemical
form; whereas HPMDI based PUs exhibited a hexagonal crystal
lattice. In addition, canola polyol-HDEDI PU demonstrated a
higher tensile strength at break than that of canola polyol-
HPMDI, attributed to the higher degree of hydrogen bonding
associated with the former sample. Nevertheless, lower
Young’s modulus and higher elongation in canola polyol-HDEDI
PU were obtained because of the flexibility of the long chain
introduced by the HDEDI diisocyanate. VC 2010 Wiley Periodicals,
1
13
structure was identified by FTIR, H NMR, C NMR, and HRMS.
This diisocyanate was used as a starting material for the prepa-
ration of entirely bio-based polyurethanes (PUs) by reacting it
with canola diol and canola polyol, respectively. The physical
properties and crystalline structure of the PUs prepared from
this diisocyanate were compared to their counterparts prepared
from similar fatty acid-derived diisocyanate, 1,7-heptamethy-
lene diisocyanate (HPMDI). The HDEDI based PUs demonstrated
various different properties compared to those of HPMDI based
PUs. For example, HDEDI based PUs exhibited a triclinic crystal
Inc. J Polym Sci Part A: Polym Chem 48: 3302–3310, 2010
KEYWORDS: biomaterials; diisocyanates from renewable sources;
metathesis; nonlinear polymers; polyurethanes; polyurethanes
from renewable sources
INTRODUCTION Polyurethanes (PUs) form a versatile class
of polymers, which are used in a broad range of applications,
for example, as elastomers, sealants, fibers, foams, coatings,
polyol. Over the past few years, our research effort has been
part of the growing worldwide interest in developing meth-
odologies for the synthesis of vegetable oil based PUs.
3
–5,7–10
1
adhesives, and biomedical materials. The synthesis of poly-
Isocyanates, the other primary building blocks of polyur-
ethanes, are normally petroleum-derived. These can be pre-
urethanes is carried out by a variety of methods, although
the most widely used method starts from di- or polyfunc-
tional hydroxycompounds (polyol) with di- or polyfunctional
isocyanates, which are usually industrially produced from
13
pared via several routes. The only route that is practiced
on a significant industrial scale is phosgenation of primary
amines or their salts. The high toxicity of phosgene, the
rather high temperature necessary to decompose the inter-
mediate carbamoyl chloride, and the poor selectivity toward
different nucleophiles are some disadvantages that limit the
synthetic use of phosgene. Some other methods which are
used on laboratory scale include the Curtius rearrangement
of acylazides, the dissociation of blocked isocyanates, the
conversion of alkyl halides with alkali cyanates, and the
dehydration of carbamic acid derivatives (adducts of amines
1
petroleum. The progressive dwindling of fossil resources,
coupled with the drastic increase in oil prices in the long-
term, have driven research to develop alternatives based on
renewable resources for the production of polymer materi-
als. In recent years, the utilization of renewable materials
such as plant oils and natural fatty acids for replacing petro-
leum derived raw materials for the production of polymeric
materials has attracted great attention, due to social, envi-
ronmental and economic considerations. In terms of bio-
based PUs, almost all current available scientific publica-
14
and carbon dioxide).
2
–11
tions
have been focused on preparing such types of poly-
In the past few decades, researchers have sought different
ways and technologies to viably produce isocyanates from
renewable resources. However, literature available regarding
the synthesis of diisocyanate from vegetable oils is very lim-
ited. The one instance of diisocyanate production from fatty
mers using polyol derived from vegetable oil in combination
with petrochemical-based diisocyanates. It has been demon-
1
2
strated that these bio-based PUs have comparable proper-
ties in many aspects with PUs derived from petrochemical
Correspondence to: S. S. Narine (E-mail: sureshnarine@trentu.ca)
Journal of Polymer Science: Part A: Polymer Chemistry, Vol. 48, 3302–3310 (2010) V
2010 Wiley Periodicals, Inc.
C
3302
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
1
5
acids is reported in a US Patent. It describes the synthesis
of diisocyanates from diamine precursors prepared from the
hydrogenation of dinitrile compounds which were obtained
from hydroxyl-substituted fatty nitrile and unsaturated
nitrile starting materials. In addition, a fatty acid-based diiso-
cyanate known as dimer acid diisocyanate containing 36 car-
bon atoms in the chain has been commercialized by Henkel
Corportion Company and General Mills. This diisocyanate has
ethyl vinyl ether, triethylamine, sodium azide, dibutyltin
dilaurate (DBTDL), Tin(II) 2-ethylhexanoate (Sn(OCT) ), an-
2
hydrous DMF, and anhydrous THF were purchased from
Sigma-Aldrich. Ethylchloroformate was obtained from BDH,
Methanol, chloroform, ethyl acetate, and hexane were pur-
chased from Fisher. Canola polyol was synthesized using an
ozonolysis and hydrogenation technology developed by our
9
,10
group—the detailed procedure was reported elsewhere.
1
6–18
been utilized for the preparation of PU coatings.
Recently, soybean oil derived isocyanate has been synthe-
sized, which involved substitution of the allylic bromides of
plant oil triacylglycerols (TAGs) with AgNCO.
More recently, a new linear saturated terminal diisocyanate,
,7-heptamethylene diisocyanate (HPMDI) has been synthe-
sized from azelaic acid via Curtius rearrangement using oleic
The chemical reaction procedure is illustrated in Scheme 1,
as an example. The hydroxyl number of the canola polyol
was 237 mg KOH/g, as determined according to the ASTM
D1957–86. The polyol contained 60.18% 6 1.16 wt %,
1
9
2
6.00% 6 0.48 wt%, and 4.72% 6 0.03 wt% of triol, diol
1
0
and mono-ol, respectively. The remainder consists of ꢀ9
wt% saturated TAGs. The canola diol was separated from
canola polyol by flash chromatography using the procedure
developed by Yue and coworkers HPMDI was synthesized
from oleic acid according to the reported procedure.
1
2
0
acid as starting materials by our research group. It has
been established that this fatty acid-derived diisocyanate
was capable of reacting with lipid-based polyol to produce
entirely bio-based PUs with comparable properties within ac-
ceptable tolerances. However, in the Curtius rearrangement
azelaic diazide intermediate can be explosive because the ra-
tio of carbon and oxygen atoms numbers (N þ N ) to nitro-
9
2
0
Fatty Acid-Derived Unsaturated Diisocyanate Synthesis
,18-Octadec-9-enedioic Acid 2
Oleic acid [28.2 g (90%), 0.1 mol] was transferred into a
1
C
O
2
4
50-mL three-necked round-bottomed flask and stirred at
5 C under nitrogen gas for 0.5 h. Grubbs catalyst 2th gen-
N
gen atoms numbers (N ) is less than 2. This can be a barrier
ꢁ
to larger scale batch type processes because extreme care is
eration (89 mg, 0.01 mmol) was then added. The reaction
required. For organic azides to be safely manipulated, (N þ
C
ꢁ
2
1
mixture was kept at 45 C and stirred with a stirrer bar. Af-
N )/N should be at least 3. To minimize this issue, we
O
N
ter ꢀ5 minutes, the diacid began to precipitate from the
reaction mixture. After 24 h, the crude product was
quenched with ethyl vinyl ether (6.5 mL), and excess ether
was removed under reduced pressure. The residue was puri-
fied by column chromatography using a hexane: ethyl acetate
eluting solvent (4:1, 3:1, 2:1, 1:1, 1:0) to give 10.0 g of pure
report on our efforts in this work as the first example of the
preparation of aliphatic diisocyanate with longer chains from
fatty acid, which could be a valuable agent for organic reac-
tions, for example, the formation of resinous materials by
reaction with glycols and diamines to from polyurethanes
and polyureas. Moreover, this report also provides a new
monomer for the production of polymers which are capable
of vulcanization or cross-linking due to the unsaturation in
the monomer unit. The introduction of C ¼¼ C double bonds
into aliphatic polymer chains would offer a starting point for
a very diverse manifold of reactivity. For example, Hartwig
and Hillmyer have reported the borylation of polybutene
and polypropylene and the oxyfunctionalization of polypro-
pylene. Furthermore, a partially unsaturated polymer could
be cross-linked with standard techniques to give materials
1
,18-octadec-9-enedioic acid 2 (Scheme 2) as a white solid
2
7
(
64%). The other option for purification is recrystallization
of the product in hexane/ethyl actetate several times.
2
2
1,16-Diisocyanatohexadec-8-ene (HDEDI) 3
2
3
A suspension of product 2 (5 g, 16 mmol) and triethylamine
(5.4 mL, 38.4 mmol) in anhydrous THF (100 mL) was cooled
2
4
ꢁ
2
5
to 0 C and ethylchloroformate (3.36 mL, 35.2 mmol) was
added dropwise. The resulting mixture was stirred for 3
ꢁ
2
6
hours at 0 C under N
2
atmosphere. Then it was added drop-
with enhanced mechanical properties.
wise to a solution of sodium azide (4.17 g, 64 mmol) in
ꢁ
To the best of our knowledge, this is the first time that a lin-
ear long chain unsaturated terminal aliphatic diisocyanate
was synthesized using a cheap and short procedure from
fatty acid. We also demonstrate the feasibility of this new
type of fatty acid-derived diisocyanate in the preparation of
entirely bio-based PUs by reacting it with canola diol and
canola polyol, respectively. It is expected that these PUs ex-
hibit various different properties to their counterparts pro-
duced from the same polyol but HPMDI.
water (70 mL) that was cooled to 0 C and kept for 1 hour,
ꢁ
followed by another hour at 5 C. The reaction mixture was
added to a separatory funnel containing 100 mL of cold
water, and the organic layer was separated and dried over
MgSO . After filtering, the solvent was completely removed
4
under reduced pressure at room temperature (because of
the long carbon chain of this diazide, it is safe under dry
conditions). Anhydrous THF (50 mL) was then added to the
flask and the resulting solution was heated to reflux for 3
hours under N2 atmosphere, then solvent was removed
under reduced pressure. 10 mL of hexane was added to the
residue and the solution was passed through a very short sil-
icagel column under N2 pressure and washed out by extra
hexane. After removing the solvent under reduced pressure,
2.8 g of diisocyanate 3 was obtained as a pale yellow oil.
EXPERIMENTAL
Materials
Oleic acid (90% purity), 1,3-Bis-(2,4,6-trimethylphenyl)-2-
imidazolidinylidene) (dichlorophenylmethylene)(tricyclohex-
ylphosphine)ruthenium (grubbs catalyst 2th generation),
(
NOVEL LONG CHAIN UNSATURATED DIISOCYANATE, HOJABRI, KONG, AND NARINE
3303

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
SCHEME 1 The chemical reac-
tion procedure of polyol from
triolein.
(
57% yield and 98% purity). Purity of compound was deter-
4 mL of anhydrous DMF and added to an addition funnel fit-
ted to the three-neck flask. The canola diol and catalyst solu-
tion was slowly (over 0.5 h) dropped into the diisocyanate
solution at room temperature. The reaction mixture was
mined by GC-FID (Fig. 1).
1
H NMR: d 1.28–1.37 (m, 16H, 8CH ), 1.59–1.62 (m, 4H,
2
2
CH CH NCO), 1.95–1.98 (m, 4H, 2CH ACH ¼¼ CH), 3.28 (t,
2
2
2
ꢁ
3
increased to 85 C and kept at this temperature for 18 hours
J
¼ 6.5 Hz, 4H, 2 OCNACH ), 5.37–5.39 (m, 2H, CH ¼¼ CH);
HH
2
1
3
with stirring. After cooling, the reaction mixture was precipi-
tated into a large excess of warm distilled water. The poly-
mer was filtered and washed with methanol. The resulting
polymer was then dissolved in 10 mL of chloroform and pre-
cipitated in 300 mL of methanol. The white solid was filtered
C NMR: d 26.5 (CH ), 28.8 (CH ), 28.9 (CH ), 29.5 (CH ),
2
2
2
2
3
1.3 (CH ), 32.5 (CH ), 43.0 (NACH ), 121.9 (NCO), 130.3
2 2 2
ꢂ1
(
CH ¼¼ CH), IR (cm ): 967, 1354, 1464 (CH2 Scissoring),
272 (N ¼¼ C ¼¼ O), 2855, 2927 (CAH). HRMS(EI) calcd for
C H O N 306.2307, found m/z 306.2305.
2
1
8 30 2 2
ꢁ
and washed by methanol and dried in vacuum oven (50 C)
Polyurethane Preparation
Polyurethanes Prepared from Canola Oil-Based Diol
for 3 days. The yield for polymerization of canola diol and
HPMDI was 76% and for canola diol and HDEDI was 74%.
The samples were coded as canola diol-HPMDI, canola diol-
HDEDI for PUs produced from canola diol with HPMDI and
HDEDI, respectively.
1
mmol of HPMDI or HDEDI was dissolved in 1 mL of anhy-
drous DMF in a stirred three-neck flask. Canola diol (ꢀ0.65
g, 1 mmol) and one drop of Sn(OCT) were dissolved in
2
SCHEME 2 Synthesis of unsaturated diisocyanates from oleic acid.
3304
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
Characterization
Gas Chromatograph-Flame Ionization Detector (GC-FID)
System
Gas chromatograms were obtained on an Agilent 6890N cap-
illary GC (Santa Clara, CA) equipped with a flame ionization
detector and Agilent 7683B auto sampler. The 30 m ꢃ 0.32
mm ꢃ 0.1 lm DB-5HT column was used for the determina-
tion of diisocyanate compound purity. The temperature of
ꢁ
the column was initially set at 85 C then increased to 250
ꢁ
ꢁ
C at a rate of 10 C/min and held for 5 min.
NMR, Mass and FTIR
1
13
H NMR and C NMR were recorded at larmor frequencies
of 500 and 125 MHz, respectively, using a Varian UNITY 500
NMR spectrometer (Varian, CA). Deuterated chloroform
(
CDCl ) was used as solvent. Mass spectra were acquired on
3
a Kratos Analytical MS-50 (Kratos Analytical, Manchester,
UK) EI high-resolution spectrometer. FTIR spectrum of diiso-
cyanate was measured with a Mattson Galaxy series 3000
FTIR spectrophotometer. The analysis of polymer specimens
was performed using a Bruker Vertex 70 FTIR main bench
with an attached Hyperion FTIR microscope using OPUS soft-
ware. The spectra were obtained using a micro-ATR objective
with an analysis area ꢀ100 lm in diameter. The spectra
were acquired using 128 scans at a resolution of four
wavenumbers.
FIGURE 1 GC-FID chromatogram of fatty acid-derived unsatu-
rated diisocyanate.
Polyurethane Prepared from Canola Oil-Based Polyol
PU sheets were prepared by reacting canola polyol with dii-
socyanates at molar ratios of the OH group to the NCO
group, that is OH/NCO of 1.0/1.1. The desired OH/NCO
molar ratio satisfies the equation:
Molecular Weight Measurements by Gel Permeation
Chromatography (GPC)
n
The number and weight-average molecular weights (M and
Wpolyol=EWpolyol
Mratio ¼ À
Á
(1)
M , respectively) were determined by GPC (Santa Clara, CA).
w
WPU ꢂ Wpolyol =EWisocyanate
The tests were carried out using GPC with an Agilent
G1311A quaternary pump, G1362A refractive index detector,
and a PL gel column (5 lm mixed-D). Chloroform and DMSO
where Wpolyol is the weight of the polyol, EWpolyol is the
equivalent weights of polyol, WPU is the total weight of PU to
produce and EWisocyanate is the equivalent weight of the
isocyanate.
(
1:1) was used as the eluent at a flow rate of 0.8 mL/min.
Sample concentrations of 0.4% (w/v) and injection volumes
of 10 lL were used. Polystyrene (PS) standards were used
to generate a calibration curve.
The equivalent weights of both HPMDI and HDEDI were cal-
culated based on their molar mass. The equivalent weight of
canola polyol was determined using the equation:
Wide Angle X-Ray Diffraction (WAXD)
A Bruker AXS X-ray diffractometer (Madison) equipped with
a filtered Cu-Ka radiation source (k ¼ 0.1542 nm) and a 2D
detector was used to record the WAXD patterns. The proce-
dure was automated and controlled by the Bruker AXS’s
molecular weight of KOH ꢃ 1000
EWpolyol ¼
OH Number
56110
¼
g per mole of hydroxyl group:
‘‘GADDs V 4.1.08’’ software. The frames were processed
OH Number
using GADDS software and the resulting spectra were ana-
lyzed using Bruker AXS’s ‘‘Topas V 2.1’’ software.
Canola polyol-HPMDI, canola polyol-HDEDI represent PUs
produced from canola oil-based polyol with HPMDI and
HDEDI, respectively.
Thermal Properties
MDSC measurements were carried out on a DSC Q100 (TA
Instruments, DE), equipped with a refrigerated cooling sys-
A suitable amount of polyol and diisocyanate were weighed
in a plastic container, stirred for 2 min before trace amounts
of DBTDL was added. The mixture was further mixed for 5
min, poured in another container and placed in a vacuum
oven at 40 C for 5 to 8 min to degas the CO released dur-
ꢁ
tem. The samples were heated at a rate of 10 C/min from
ꢁ
ꢁ
25 C to 100 C to erase thermal history, then cooled down
ꢁ
ꢁ
to –50 C at a cooling rate of 5 C/min. MDSC measurements
were performed with a modulation amplitude of 1 C/min
and a modulation period of 60 s at a heating rate of 3 C/
min to 100 C. The second heating stage was selected for
ꢁ
ꢁ
2
ꢁ
ing the side reaction of diisocyanate with moisture or car-
boxylic acid and the air trapped during mixing. Air was then
introduced to the oven to avoid the deformation of the sam-
ple under vacuum and the sample was postcured for about
ꢁ
the analyzing of heating data. All the DSC measurements
were performed following the ASTM E1356–03 standard
procedure under a dry nitrogen gas atmosphere.
ꢁ
ꢁ
2
4 hours at 70 C, 10 hours at 110 C.
NOVEL LONG CHAIN UNSATURATED DIISOCYANATE, HOJABRI, KONG, AND NARINE
3305

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
FIGURE 2 HRMS spectrum of fatty acid-derived unsaturated
diisocyanate.
FIGURE 3 FTIR spectra of (a) fatty acid-derived unsaturated dii-
socyanate, (b) canola diol-HPMDI, (c) canola diol-HDEDI, and
(d) canola polyol-HDEDI polyurethane.
Mechanical Properties
Mechanical properties were tested using an Instron (MA)
tensile testing machine (model 4202) equipped with a 50
Kgf load cell and activated grips which prevented slippage of
the sample before break. Specimens were cut out from the
PUs sheets using an ASTM D638 Type V cutter. The measure-
ments were performed at room temperature with cross-head
speed of 50 mm/min, as suggested by the aforementioned
ASTM standard. The data presented were average of five dif-
ferent measurements. The reported errors are the subse-
quent standard deviations.
1
3
of double bond. Also C NMR showed a signal at d121.9
and 130.3, which corresponds to the isocyanate group and
double bonded carbons. High resolution mass spectrum (Fig.
2) of the diisocyanate 3 was in accordance with the struc-
ture proposed. The chemical structure of the diisocyanate
obtained was further confirmed by FTIR. As shown in Figure
ꢂ1
3 (curve a), a strong characteristic band at 2274 cm which
corresponds to the N ¼¼ C ¼¼ O stretching vibration was
observed.
RESULTS AND DISCUSSION
Characterization of Diisocyanate
Canola Diol-Based Polyurethane
The self-metathesis of readily available monounsaturated
fatty acids has the potential of being an important pathway
for the synthesis of symmetrical long-chain unsaturated
dicarboxylic acids. By using Grubbs catalyst and the self-me-
tathesis procedure, unsaturated dicarboxylicacid 2 (Scheme
In addition to the presence of C ¼¼ C double bond, the number
of methylene groups of the synthesized HDEDI diisocyanate
monomer is 14, which will offer different characteristics
when it is compared with HPMDI (whose methylene groups
number is 7) synthesized previously in our group. Moreover,
all of the current available vegetable oils-based polyol,
including canola polyol, is a mixture of triol, diol and mono-
2
7
2) was synthesized. It has been reported that the double
bond on the alkyl chain of diacid 2 is in a trans geometry,
2
7
10
determined from IR spectrum. Unsaturated diacylazide was
ol due to the nature of vegetable oils feedstock. To synthe-
2
0
prepared from the diacid 2 via the Curtius reaction route.
size model renewable resources based PUs whose purity and
structure could be ensured, canola diol was separated from
the mixture of canola polyol by flash chromatography using
The white solid diacyl azide was dissolved in anhydrous THF
and decomposed by heating at reflux to give the correspond-
ing pale yellow diisocyanate 3 that was further purified by
passing through a very short chromatograghy column of sili-
9
the procedure developed by Yue and coworkers and its
structure was illustrated in Scheme 1. The obtained canola
diol was utilized to react with HDEDI and HPMDI to prepare
polyurethanes, respectively. The number average molecular
1
cagel. The H NMR spectra of the diisocyanate 3 showed a
triplet at d3.28, related to the CH2 group attached to
N ¼¼ C ¼¼ O and a multiplet at d5.38, related to the CAH groups
4
weights are 4.0 ꢃ 10 (PDI ¼ 1.5) for canola diol-HDEDI PU
TABLE 1 Average Molecular Weight Determined by GPC and Thermal Properties of Polyurethane Made from Different Raw
Materials Determined by MDSC
ꢁ
ꢁ
M
w
M
n
PDI
T
g
( C)
T
m
( C)
4
4
4
4
Canola diol-HDEDI
Canola diol-HPMDI
Canola polyol-HDEDI
Canola polyol-HPMDI
6.0 ꢃ 10
5.0 ꢃ 10
4.0 ꢃ 10
3.6 ꢃ 10
1.5
1.4
80
73
48
62
ꢂ23
ꢂ16
3306
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
ꢁ
peaks were observed. One is in the range of 20–25 C, the
ꢁ
other one is in the range of 70–80 C. It has been
reported
2
9–31
that the peak in the low temperature range
can be ascribed to the disruption of urethane-polyol bonds,
while peaks in the higher temperature range (usually
assigned to the melting point of the polymers) was ascribed
to the disruption of urethane-urethane bonds. Additionally,
it should be noted that the melting point (T ) of canola
m
ꢁ
diol-HDEDI PU is 80 C, while that of canola diol-HPMDI is
C (Table 1). Both Tm values are much lower than that
of the PUs (ca. 140 C) produced from 1,6-hexamethylene
ꢁ
73
ꢁ
3
2
diisocyanate (HDI) and long-chain, aliphatic a,x-diol. This
was caused by the softening effect of the longer HDEDI
chains on the crystal structures as well as the defects in
the crystals of canola diol-based PUs due to the presence of
dangling chains in canola diol as illustrated in Scheme 1.
These long saturated fatty acid branches in canola-diol mol-
ecules interrupt the crystal packing, therefore resulting in a
less effectively packed and less stable structure. Addition-
ally, from a thermodynamic point of view, the decreasing of
FIGURE 4 MDSC curves of canola diol-HPMDI, and canola diol-
HDEDI polyurethane.
T
m
in this study is due to the increasing of the entropic
4
and 3.6 ꢃ 10 (PDI ¼ 1.4) for canola diol-HPMDI PU deter-
effect in the highly branched canola diol-based PUs. It is
also worth mentioning that no glass transition was
observed in these two PU samples. This might be due to
the fact that heat capacity changes over the glass transition
region of these types of materials are too small to be
mined by GPC (Table 1). It is worthmentioning that the num-
ber-average molecular weight values determined by GPC are
relative to those of PS, however, the chemical structure dif-
ference between TPU and PS might cause errors, therefore,
further work needs to be done to obtain reliable average
molecular weights by utilizing other techniques such as
vapor phase osmometry, light scattering and so forth.
detected or the T ’s were overlapped by the endothermic
g
peaks at low temperatures.
The WAXD patterns of canola diol-HPMDI and canola diol-
HDEDI PUs are shown in Figure 5. Canola diol-HDEDI PU
exhibited three peaks with d-spacings of 1.45 nm, 0.43 nm
and 0.39 nm, which correspond to the (004), (100), and
(010) planes of typical aliphatic polyurethanes in the usual
The FTIR spectra for canola diol-HPMDI and canola diol-
HDEDI PUs are shown in Figure 3 (curves b and c, respec-
tively). It is clear that in both specimens, the N ¼¼ C ¼¼ O
stretching vibration band was missing, meanwhile, a strong
ꢂ1
33
stretching band located around 3330 cm characteristic of
triclinic crystal phase. This triclinic crystal form consists of
the NAH group, and stretching vibration bands characteris-
tic of free nonhydrogen bonded C ¼¼ O groups (ca. 1740
hydrogen-bonded sheets stacking via van der Waals forces,
and the interchain hydrogen bonds are confined to the plane.
The length of 1.45 nm of the chain axis repeat distance is
ꢂ1
cm ) and hydrogen bonded urethane C ¼¼ O groups in or-
ꢂ1
dered (crystalline) domains (ca. 1685 cm ) were present
in the FTIR spectra. These results indicates that this novel
fatty acid-derived unsaturated diisocyanate was successfully
reacted with canola diol produced PUs. Nevertheless, it is
worth pointing out that the wave number of NAH stretch-
ꢂ1
ing band of canola diol-HDEDI PU is 3327 cm
and that
ꢂ1
of canola diol-HPMDI PU is 3336 cm . Because higher
bond strength is generally characterized by lower wave
2
8
numbers,
this result suggests that the hydrogen bond
strength for the former specimen is higher than the latter
one, even though the higher number of methylene groups
in HDEDI decreases the concentration of urethane groups
and hydrogen bond strength. This difference is mainly
attributed to the fact that a fully saturated and optimum
hydrogen bonding scheme with all trans conformation in
the entire polymer backbone could be active in the former
specimen.
The reversing heat flow versus temperature curves obtained
for canola diol-HPMDI and canola diol-HDEDI PUs samples
are shown in Figure 4. In both cases, two endothermic
FIGURE 5 Wide angle X-ray diffraction patterns of canola diol-
HPMDI, and canola diol-HDEDI polyurethane.
NOVEL LONG CHAIN UNSATURATED DIISOCYANATE, HOJABRI, KONG, AND NARINE
3307

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
The crystal structures of canola polyol-HDEDI and canola
polyol-HPMDI PU are observed by WAXD as well (data not
shown). Similar to the corresponding canola diol-based PUs,
canola polyol-HDEDI PU exhibited three peaks with d-spac-
ings of 1.41 nm, 0.44 nm, and 0.39 nm, and canola polyol-
HPMDI exhibited two peaks with d-spacings of 1.10 nm and
0.42 nm. This result confirms that the diols in the canola
polyol reacted with diisocyanate and formed crystals with
similar structures to corresponding canola diol-based PUs.
Nevertheless, the crystalline peaks in canola polyol-based
PUs are broader which is due to the disturbance to the pack-
ing of the crystals due to the presence of mono-ols and satu-
rated TAGs.
The reversing heat flow versus temperature curves obtained
for both canola polyol-based PU samples are shown in Fig-
ure 7. T s are observed in both cases, which complement
m
the FTIR and WAXD results. However, the low temperature
endothermic peaks which occurred in canola diol-based PUs
are missing, which might be due to the interruption of the
complex structure of canola polyol-based PU. Meanwhile, in
contrast to canola diol-based PUs, the Tm of canola polyol-
HDEDI PU is lower than that of canola polyol-HPMDI PU.
This probably results from the fact that canola polyol-
HDEDI PU contains substantially longer segments which
provide more spaces for inactive molecules (mono-ol and
saturated TAG) to entrap between well packed molecules.
Those entrapped molecules act as defects interrupting the
crystal packing and causing a less well packed and less sta-
ble structure. In addition, the glass transition temperature
FIGURE 6 A representative spectrum in the carbonyl region for
the canola polyol-HDEDI polyurethane, along with the individu-
ally fitted peaks from the deconvolution procedure.
slightly higher than the theoretical value, 1.39 nm, calculated
for planar zigzag chains with the accepted bond angles and
bond lengths. This is due to crystal structural distortion
aroused from the presence of the double bond in HDEDI and
the long branches in the canola diol. Canola diol-HPMDI PU
exhibited two peaks in the WAXD patterns with d-spacings
of 1.08 nm and 0.43 nm, which correspond to the (004) and
3
4–36
(
100) planes of the hexagonal lattice.
In this case, the
chain axis is orthogonal to the lamellar surface with the
hydrogen bonds built up within a crystal plane with a length
of 0.43 nm.
(
Tg) was determined from Figure 7 and the results were
listed in Table 1. As expected, the T of canola polyol-HDEDI
g
ꢁ
is about 7 C lower than that of canola polyol-HPMDI,
The above results demonstrate that the synthesized fatty
acid derived unsaturated diisocyanate is capable of produc-
ing entirely bio-based polyurethanes by reacting with canola
oil derived diol. However, the separation of diol from polyol
mixture consumes a large amount of solvent and time which
would limit its application at an industrial scale. Therefore,
it is necessary to study bio-based polyurethanes prepared
from this fatty acid derived unsaturated diisocyanate and
crude canola polyol.
mainly because of the increasing flexibility caused by
the long chain of HDEDI. Moreover, the presence of C ¼¼ C
double bonds in HDEDI molecules also contributes to the
decreasing of T because there is little steric hindrance
g
adjacent to this bond.
Canola Polyol-Based Polyurethane
The FTIR spectrum for canola polyol-HDEDI (curve d, Fig. 3)
clearly demonstrates that the N ¼¼ C ¼¼ O stretching vibration
band is missing, meanwhile, a strong stretching band located
ꢂ1
at 3330 cm characteristic of the NAH group and stretch-
ꢂ1
ing vibration bands around 1700 cm characteristic of the
C ¼¼ O group were present. Figure 6 is an enlarged FTIR spec-
trum in the carbonyl region along with the individual fitted
bands from a deconvolution procedure performed on the
spectrum. The four bands in the spectrum are identified as
follows: free non hydrogen bonded C ¼¼ O groups (1741
ꢂ1
cm ), hydrogen bonded urethane C ¼¼ O groups in disordered
ꢂ1
(
amorphous) domains (1721 cm ), hydrogen bonded ure-
thane C ¼¼ O groups in ordered (crystalline) domains (1690
ꢂ1
cm ), and the C ¼¼ C double bond group in the diisocyanate
FIGURE 7 MDSC curves of canola polyol-HPMDI, and canola
ꢂ1
chains (1653 cm ).
polyol-HDEDI polyurethane.
3308
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA

ARTICLE
of the PUs are greatly dependent on the intermolecular
hydrogen bonding because the increased intermolecular
3
7
hydrogen bonding improves the aggregating strength.
When HDEDI with longer molecular chains was used in the
PU preparation, despite the decrease in the concentration of
urethane groups, the overall hydrogen bond strength is still
higher. Therefore, the canola polyol-HDEDI samples demon-
strated a higher tensile strength at break.
CONCLUSIONS
A novel long chain linear unsaturated terminal diisocyanate
has been successfully synthesized from oleic acids via Cur-
tius rearrangement. The feasibility of utilizing this new diiso-
cyanate for the production of bio-based polyurethanes has
been demonstrated by reacting it with canola diol and canola
polyol, respectively. The crystalline structure and physical
properties of these polyurethanes were compared to their
counterparts made from a similar fatty acid-derived diisocya-
nate, 1,7-heptamethylene diisocyanate. For model PUs pro-
duced from canola diol, the overall hydrogen bond strength
of canola diol-HDEDI PU is higher than that of canola diol-
HPMDI PU. In addition, both HDEDI based PUs exhibit the
triclinic crystal subcell, whereas HPMDI based PUs exhibit a
hexagonal crystal lattice subcell. Because of the flexibility of
the long chain introduced by the HDEDI diisocyanate, lower
Young’s modulus and higher elongation in canola polyol-
HDEDI PU were observed. On the other hand, canola polyol-
HDEDI PU demonstrated a higher tensile strength at break
than that of canola polyol-HPMDI, which is attributed to the
higher intermolecular forces (hydrogen bonding) associated
with the former sample.
FIGURE 8 Nominal stress versus strain curves of canola polyol-
HPMDI, and canola polyol-HDEDI polyurethane.
It is worth mentioning that the formation of vegetable oil-
based PU networks (from three or more reactants) is more
complicated because of the nature of the vegetable oil feed-
stock. In the case of the canola polyol-based PUs studied in
this work, since the starting materials, that is canola polyol,
is a mixture of triol, diol and mono-ol, the relative reactivity
of functional groups are not equal, which results in structur-
ally inhomogeneous systems containing various domains
with different sizes. Each domain consists of alternate polyol
(
triol or diol)-isocyanate sequences. Part of the domain is
chemically cross-linked through covalent bonds between dii-
socyanate with triol, diol and mono-ol, and part of it is crys-
talline domain resulting from physical cross-linking (hydro-
gen bonding) between segments. The packing of these
crystals was interrupted by the mono-ol, saturated TAG and
the dangling chains in the diol molecules. These chains
would restrict the effective packing of the polymer chains
into crystals by creating steric hindrances. Additionally, the
chemical cross-links would reduce the mobility of molecular
chains and cause a reduction of these chains to orient to
form crystals.
This work was supported by Bunge Oils, NSERC, Alberta Canola
Producers Commission, Alberta Agricultural Research Institute,
and Alberta Crop Industry Development Fund. The authors
thank Mr. Ereddad Kharraz and Mr. Ali Mahdevari for technical
support.
REFERENCES AND NOTES
1
Szycher, M. Szycher’s Handbook of Polyurethanes; CRC
Press: Boca Raton, FL, 1999.
The stress versus strain curves for canola polyol-HDEDI and
canola polyol-HPMDI samples are shown in Figure 8. For
canola polyol-HPMDI PU, the tensile strength at break is 15
2
Heidbreder, A.; H o¨ fer, R.; Gr u¨ tzmacher, R.; Westfechtel, A.;
Blewett, C. W. Fett/Lipid 1999, 101, 418–424.
3
Kong, X.; Narine, S. S. Biomacromolecules 2007, 8,
203–2209.
6
2 MPa, Young’s modulus is 92 6 3 MPa and maximum
2
elongation is 133% 6 13%, whereas for canola polyol-
HDEDI PU, the tensile strength at break is 21 6 2 MPa,
Young’s modulus is 72 6 4 MPa and maximum elongation is
4 Kong, X.; Yue, J.; Narine, S. S. Biomacromolecules 2007, 8,
3584–3589.
1
75% 6 22%. The decreases in Young’s modulus and
5
Kong, X. H.; Narine, S. S. Biomacromolecules 2008, 9,
424–1433.
increases in maximum elongation in canola polyol-HDEDI PU
are due to the flexibility of the long chain introduced by the
HDEDI diisocyanate. Additionally, the canola polyol-HDEDI
sample demonstrated a higher tensile strength at break than
that of canola polyol-HPMDI. This behavior is attributed to
the higher intermolecular forces associated with the former
sample, since fully saturated hydrogen bonding in the entire
polymer backbone was generated. The mechanical properties
1
6
Lu, Y. S.; Larock, R. C. Biomacromolecules 2007, 8,
108–3114.
3
7
Narine, S. S.; Kong, X.; Bouzidi, L.; Sporns, P. J Am Oil
Chem Soc 2007, 84, 55–63.
8 Narine, S. S.; Kong, X.; Bouzidi, L.; Sporns, P. J Am Oil
Chem Soc 2007, 84, 65–72.
NOVEL LONG CHAIN UNSATURATED DIISOCYANATE, HOJABRI, KONG, AND NARINE
3309

JOURNAL OF POLYMER SCIENCE: PART A: POLYMER CHEMISTRY DOI 10.1002/POLA
9
Narine, S. S.; Yue, J.; Kong, X. J Am Oil Chem Soc 2007, 84,
23 Kondo, Y.; Garcia-Cuadrado, D.; Hartwig, J. F.; Boaen, N. K.; Wag-
173–179.
ner, N. L.; Hillmyer, M. A. J Am Chem Soc 2002, 124, 1164–1165.
10 Yue, J.; Narine, S. S. J Am Oil Chem Soc 2007, 84, 803–807.
24 Bae, C.; Hartwig, J. F.; Harris, N. K. B.; Long, R. O.; Ander-
son, K. S.; Hillmyer, M. A. J Am Chem Soc 2005, 127, 767–776.
11 Zlatanic, A.; Lava, C.; Zhang, W.; Petrovic, Z. S. J Polym Sci
Part B: Polym Phys 2004, 42, 809–819.
25 Boaen, N. K.; Hillmyer, M. A. Macromolecules 2003, 36,
7
027–7034.
12 Javni, I.; Zhang, W.; Petrovic, Z. S. J Appl Polym Sci 2003,
8
8, 2912–2916.
26 Ho, T.; Martin, J. M. Structure, Properties and Applications
of Polyolefin Elastomers Produced by Constrained Geometry
Catalysts; Wiley: New York, 2000.
13 Ulrich, H. Chemistry and Technology of Isocyanates; Wiley:
Chichester, 1996.
2
7 Ngo, H. L.; Jones, K.; Foglia, T. A. J Am Oil Chem Soc 2006,
1
4 Waldman, T. E.; McGhee, W. D. Chem Commun 1994, 8,
8
3, 629–634.
957–958.
2
8 Petrovic, Z. S.; Ferguson, J. Prog Polym Sci 1991, 16, 695–836.
15 Kamal, M. R.; Dhahran, S. A.; Kuder, R. C. U.S. Pat.
3,691,225, 1972.
29 Seymour, R. W.; Cooper, S. L. Macromolecules 1973, 6, 48–53.
16 Coady, C. J.; Krajewski, J. J.; Bishop, T. E. U.S. Pat.
30 Seymour, R. W.; Cooper, S. L. J Polym Sci Part B: Polym
4,608,409, 1986.
Lett 1971, 9, 689–694.
17 Bishop, T. E.; Coady, C. J.; Zimmerman, J. M. U.S. Pat.
31 Clough, S. B.; Schneider, N. S. J Macromol Sci Phys 1968,
4,609,718, 1986.
2, 553–566.
18 Cacella, A. F. U.S. Pat. 3,622,435, 1971.
32 McKiernan, R. L.; Gido, S. P.; Penelle, J. Polymer 2002, 43,
3
007–3017.
19 Cayli, G.; Kusefoglu, S. J Appl Polym Sci 2008, 109,
2948–2955.
33 McKiernan, R. L.; Sikorski, P.; Atkins, E. D. T.; Gido, S. P.;
Penelle, J. Macromolecules 2002, 35, 8433–8439.
20 Hojabri, L.; Kong, X.; Narine, S. S. Biomacromolecules 2009,
1
0, 884–891.
34 Kinoshita, Y. Makromol Chem 1959, 33, 1–20.
35 Kinoshita, Y. Makromol Chem 1959, 33, 21–31.
21 Br a¨ se, S.; Friedrich, A.; Gartner, M. Synthesis of Hetero-
cycles via Cycloadditions I; Springer: Berlin/Heidelberg, 2008;
Vol. 12, pp 45–115.
3
6 Cui, X. W.; Li, W. H.; Yan, D. Y. Polym Eng Sci 2005, 45,
1
673–1679.
2
2 Ray, A.; Zhu, K. M.; Kissin, Y. V.; Cherian, A. E.; Coates, G.
3
7 Zia, K. M.; Barikani, M.; Bhatti, I. A.; Zuber, M.; Bhatti, H. N.
W.; Goldman, A. S. Chem Commun 2005, 3388–3390.
J Appl Polym Sci 2008, 109, 1840–1849.
3310
INTERSCIENCE.WILEY.COM/JOURNAL/JPOLA