Solubility in CO2 and carbonation studies of epoxidized fatty acid diesters: Towards novel precursors for polyurethane synthesis

Article abstract of DOI:10.1039/c0gc00371a

Novel linear polyurethanes were synthesized by bulk polyaddition of diamines with two vegetable-based biscarbonates produced from oleic acid methyl ester. Internal carbonated fatty acid diester (ICFAD) and terminal carbonated fatty acid diester (TCFAD) were obtained by the reaction of their epoxide precursors with CO₂. Terminal epoxy fatty acid diester (TEFAD) was found to be more soluble and more reactive in CO₂ than internal epoxy fatty acid diester (IEFAD). Polyurethanes obtained by polyaddition of TCFAD and ICFAD with diamines exhibit molecular weights up to 13500 g mol^-1 and glass transitions around -15 °C. Amide linkages were not observed when secondary diamine was used as the comonomer.

Full text of DOI:10.1039/c0gc00371a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/greenchem | Green Chemistry
Solubility in CO and carbonation studies of epoxidized fatty acid diesters:
2
towards novel precursors for polyurethane synthesis†
Aur e´ lie Boyer,a,b Eric Cloutet,a,b Thierry Tassaing, Benoit Gadenne, Carine Alfos and Henri Cramail*
c
d
d
a,b
Received 27th July 2010, Accepted 5th October 2010
DOI: 10.1039/c0gc00371a
Novel linear polyurethanes were synthesized by bulk polyaddition of diamines with two
vegetable-based biscarbonates produced from oleic acid methyl ester. Internal carbonated fatty
acid diester (ICFAD) and terminal carbonated fatty acid diester (TCFAD) were obtained by the
reaction of their epoxide precursors with CO
found to be more soluble and more reactive in CO
IEFAD). Polyurethanes obtained by polyaddition of TCFAD and ICFAD with diamines exhibit
2
. Terminal epoxy fatty acid diester (TEFAD) was
2
than internal epoxy fatty acid diester
(
-
1
◦
molecular weights up to 13 500 g mol and glass transitions around -15 C. Amide linkages were
not observed when secondary diamine was used as the comonomer.
Introduction
phosgene, remains an important issue to overcome. Therefore,
the development of alternative methods for the preparation
of PUs which avoid the use of phosgene is of high interest.
Recently, Narine and co-workers investigated the production of
a new linear saturated terminal diisocyanate synthesized from
With an annual world production of about 10 wt%,
polyurethanes (PUs) represent an important class of polymer
materials, widely used in many fields either as cross-linked
materials or as thermoplastics (TPU). The latter can be obtained
as highly resilient flexible foams, durable elastomeric materials,
15
oleic acid. Cramail and co-workers have also recently reported
the synthesis of polyurethanes from oleic and ricinoleic acids
using an AB-type self-condensation polymerization route from
1
high performance adhesives and sealants, medical devices, etc.
Most PUs are generally obtained from the tin-catalyzed polyad-
dition reaction between synthetic polyols and polyisocyanates.
Due to the uncertain cost of petroleum in the near future as
well as the need to develop sustainable chemical routes, the use
of renewable resources such as vegetable oils and fats for the
synthesis of polymers is currently expanding. Many types of
vegetable oil such as naturally hydroxylated ricinoleic oil and
also rapeseed oil, tung oil, linseed oil, canola oil, sunflower oil
and soybean oil have already been reported as sources of polyols
for PU synthesis. These vegetable-based polyols have been found
to be good substitutes to synthetic ones, giving rise to PUs with
versatile and competitive thermo-mechanical properties.
16
a-hydroxyl, w-acyl-azido precursors. Another non-isocyanate
route to PUs involves the reaction of polyamines with poly-
carbonates, the latter being obtained by the reaction of polye-
poxides with CO , leading to non-isocyanate polyurethanes
2
17–20
(
NIPUs).
An important benefit of this chemical pathway
, an abundant, renewable and
environmentally friendly chemical.
2–5
is to substitute phosgene by CO
2
Epoxidized vegetable oils (triglycerides) are good raw ma-
terials for carbonation. Wilkes et al. converted epoxidized
soybean oil (ESBO) into carbonated soybean oil (CSBO) in
◦
7
0 h at atmospheric pressure and 110 C, in the presence of
21
tetrabutylammonium bromide (TBABr) as the catalyst. The
authors also investigated the synthesis of NIPU networks by
reacting CSBO with different diamines. The yield and kinetics
The chemical modification of unsaturated vegetable oils
(
triglycerides) to synthesize polyols is described in the literature.
6
–9
10,11
Hydroxylation or hydroformylation
ozonolysis and transesterification
of the double bonds,
are the main reactions
of carbonation was later improved by Doll et al. thanks to the
12
13,14
22,23
use of supercritical carbon dioxide, scCO
2
.
Rokicki and co-
studied. Despite the interest in developing new vegetable-based
polyols, the use of toxic polyisocyanates, manufactured from
workers also used CSBO to modify a bisphenol-A based epoxy
24
resin. The effect of CSBO content on the curing behaviors and
the thermal and mechanical interfacial properties of the epoxy
resin/CSBO composition was analyzed.
a
Universit e´ de Bordeaux, Laboratoire de Chimie des Polym e` res
Organiques, UMR 5629, IPB/ENSCBP, 16 avenue Pey-Berland,
F-33607, Pessac cedex, France. E-mail: cramail@enscbp.fr;
Fax: (+33)5 4000 8487; Tel: (+33)5 4000 6254
Literature data mainly relates to PU synthesis from car-
bonated triglycerides. The high functionality and the varied
molecular structures of these carbonated triglycerides give
rise to rather ill-defined cross-linked PU materials. All these
considerations prompted us to investigate the use of well-defined
bis-carbonated vegetable-based precursors, for the synthesis of
linear polyurethanes (TPUs). To that purpose, we have designed
two new epoxidized fatty acid diesters (EFADs) (see Scheme
b
Centre National de la Recherche Scientifique, Laboratoire de Chimie des
Polym e` res Organiques, F-33607, Pessac cedex, France
c
Universit e´ de Bordeaux, Institut des Sciences Mol e´ culaires, UMR 5255,
3
51, Cours de la Lib e´ ration, F-33405, Talence Cedex, France
d
ITERG, 11 rue Gaspard Monge, Parc Industriel, Pessac Cedex, F
3
3600, France
†
2
Electronic supplementary information (ESI) available: CO sorption
1
), terminal and internal epoxidized fatty acid diesters (TEFAD
in the EFAD phase as a function of pressure at different temperatures.
See DOI: 10.1039/c0gc00371a
and IEFAD), respectively.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 2205–2213 | 2205

View Article Online
steel cell equipped with four cylindrical sapphire windows,
two windows for the infrared absorption measurements with
a path length of 7.2 mm and two sapphire windows for direct
33
observation of the solution. Using such windows, one can
-
1
access the wavenumber range from 2500 to 7500 cm . For
the kinetic study of the carbonation reaction, we have used
germanium windows with a path length of 100 mm instead
of sapphire ones in order to measure the infrared spectra in
-
1
the wavenumber range extending from 700 to 5000 cm . The
windows were positioned on the flat surface of the plug. A
ꢀ
R
1
00 mm Kapton foil placed between the window and the plug
compensated for any imperfections between the two surfaces.
ꢀ
R
Flat Teflon seals insured a seal between the plug and the cell
body. Heating was achieved by using four cartridge heaters
distributed throughout the body of the cell in which two
thermocouples were placed. The first one located close to one
cartridge is used for the temperature regulation and the second
was kept close to the sample area in order to measure the
temperature of the sample with an accuracy of about DT ~
◦
±
0.5 C. The cell was connected via a stainless steel capillary to
a hydraulic pressurizing system which allows the pressure to be
raised up to 50 MPa with an absolute uncertainty of ±0.1 MPa
and a relative error of ±0.3%.
The kinetics of polyurethane synthesis from the polyaddition
of carbonated fatty acid diesters (CFADs) with the two diamines
Scheme 1 Reaction scheme of IEFAD and TEFAD synthesis.
In the first part of this study, we investigate (through in situ
FTIR monitoring) the mutual miscibility of CO and the two
2
EFADs as well as the optimal conditions for the carbonation
reaction in the presence of TBABr as the catalyst, with respect
(EDA and IPDA) were carried out by means of in situ ATR-
FTIR spectroscopy by using a ThermoOptek 6700 spectrometer
and a heatable Golden Gate diamond ATR accessory (Specac).
The CFAD was mixed with the diamine, and a small droplet of
the mixture was deposited 5 minutes later on the ATR crystal,
which was preheated to the appropriate temperature. Infrared
to temperature and CO pressure. It is worth noting that
2
thermodynamic models and interaction parameter values of
different vegetable oils in supercritical carbon dioxide have
25–31
already been reported
but, to our knowledge, never for
epoxidized oils. In the second part of the study, the synthesis of
polyurethanes from the so-formed carbonated fatty acid diesters
CFAD) and two diamines (ethylenediamine and isophorone
-1
spectra with a spectral resolution of 4 cm were taken every
10 min over a period of 24 h.
(
Size exclusion chromatography (SEC) analyses were per-
diamine) is discussed. In situ ATR-FTIR spectroscopy was
found to be a useful technique to follow the formation of
polyurethanes as a function of temperature and the nature of
the diamines.
formed on polyurethane samples before precipitation, at room
temperature, in THF with a setup consisting of a WATERS
8
80-PU pump and a series of three microstyragel columns
˚
with pore sizes of 103, 105 and 106 A. The eluate of the
filtered samples was monitored using simultaneous UV and
refractive index detections. The elution times were converted
to molecular weight using a calibration curve based on low
Experimental section
Materials
dispersity (M
w
/M ) polystyrene (PS) standards. The differential
n
scanning calorimetry (DSC) thermograms were measured using
a DSC Q100 apparatus from TA instruments. Polymer samples
were first heated from -100 to 200 C and the glass transition
IEFAD and TEFAD precursors have been synthesized from
32
oleic fatty acid methyl ester from sunflower oil (Scheme 1).
◦
1
The purity of each EFAD, determined by H NMR was close to
5%. Tetrabutylammonium bromide (TBABr), ethylenediamine
EDA) and isophorone diamine (IPDA) were purchased from
Sigma Aldrich and used as received. Carbon dioxide N45 (purity
9,95%) was supplied by Air Liquide.
temperatures were determined after the second heating run. All
9
(
◦
-1
runs were performed at a heating rate of 10 C min .
9
Experimental procedure
In order to determine the EFAD concentration in the CO -
2
Analysis
rich phase, the lower part of the cell was first partially filled
with EFAD in order to get its level well below the incoming
infrared beam (see Fig. 1a). For the determination of the EFAD
The IR measurements were performed on a Thermo-Optek
interferometer (type 6700) equipped with a globar source, a
Ge/KBr beam splitter and a DTGS detector. Single beam
and CO
temperature and CO
2
concentration in the oil-rich phase as a function of
pressure, the cell was filled to a level
-
1
spectra recorded in the spectral range 400–7500 cm with a
2
1
2
cm- resolution were obtained by Fourier transformation
above the incoming beam (see Fig. 1b). In each case, the cell
was then heated up to the required temperature. The spectra
of 50 accumulated interferograms. The in situ phase behavior
investigations were performed using a home-made stainless
were recorded for the raw EFAD, then CO
2
was added up to the
2
206 | Green Chem., 2010, 12, 2205–2213
This journal is © The Royal Society of Chemistry 2010

View Article Online
Fig. 1 Schematic drawing of the in situ experiment using a transmission
optical cell to study a) the concentration of EFAD in CO phase and b)
the CO sorption and EFAD swelling.
2
2
highest desired pressure. The system was kept under isobaric and
isothermal conditions for about 15 min in order to ensure that
the equilibrium was achieved. The stabilization of the operating
conditions was controlled by recording consecutive spectra. The
mixture was constantly homogenized during the experiment
using a magnetically driven stirrer positioned at the bottom of
the cell.
For the kinetic study, the experimental procedure was almost
the same except that the catalyst (TBABr, 3 wt%) was first
dissolved in the EFAD-rich phase and the in situ FTIR
Fig. 3 IR spectra of IEFAD in the oil-rich phase at 7.5 MPa (black),
◦
1
5 MPa (red) and 20 MPa (blue), at 70 C.
wavenumber range 4600–6000 cm- that occur with an increase
1
of CO pressure.
2
A number of significant peaks associated to the combination
or overtones of EFAD can be observed in the spectral range
monitoring was started immediately after the injection of CO
2
-
1
at the selected pressure and temperature. One infrared spectrum
was recorded every 10 min.
5200–6000 cm . Increasing the CO
2
pressure leads to a decrease
in the intensities of the EFAD bands (for example at 5690 and
-
1
5
800 cm ), whereas the intensity of the CO
of CO sorbed into EFAD (for example at 4950 cm ), increases.
As the more intense peak observed in Fig. 3 centred at
2
peaks, characteristic
-
1
2
Results and discussion
-
1
Phase behaviour of the EFAD-CO
2
system
5800 cm is sometimes saturated in our experimental conditions,
-
1
the peak centred at 5690 cm was used to determine the
concentration of the EFAD subjected to scCO
Concerning CO , one can detect three peaks at 4800, 4950 and
100 cm which are assigned to the combination modes 4n +n
Infrared absorption spectra. The infrared spectra of both
2
.
EFADs (IEFAD and TEFAD) and the CO -rich phases of the
2
2
EFAD/CO
2
mixtures were recorded for a range of pressures up
-
1
5
n
2
4
3
,
◦
to 20 MPa at temperatures from 25 to 120 C. Fig. 2 illustrates
3
1
+2n
2
+n
3
, and 2n
1
+n
3
of the CO molecule respectively. Due
2
(
for the case of IEFAD) the spectral changes in the CO
phase that occur with an increase of CO pressure. A number of
significant peaks associated with combination bands of CO and
fundamental stretching modes of EFAD can be observed. The
intensities of these peaks increase with CO pressure indicating
an increase of the CO density and the EFAD solubility in the
CO -rich phase. The peak centered at 2860 cm , associated with
the CH stretching vibrations of EFAD, can be used to determine
the solubility of EFAD in CO . In the same manner, Fig. 3
illustrates the spectral changes of the IEFAD-rich phase in the
2
-rich
-
1
to the weak intensity of the peaks at 4800 and 5100 cm , the
2
-
1
band at 4950 cm was used to estimate the variation of the
CO concentration incorporated into the EFAD-rich phase as a
function of the CO pressure.
2
2
2
2
2
Data processing for the determination of the mutual solubility
-
1
2
of EFAD and CO2. In order to determine the concentration of
-
1
EFAD in the CO -rich phase, the peak centered at 2860 cm
2
2
associated with CH stretching vibrations of EFAD was selected.
The peak height of these characteristic infrared bands allows
us to determine the concentration of the solute according to
the Beer–Lambert law, A = e. l. c where A is the sample
-
1
-1
absorbance; e the molar extinction coefficient (L mol cm );
l, the optical path length (cm) and c the sample concentration
-
1
(
mol L ). In this study, the peak height was used instead of
integrated area because of the superposition of another CH
band in the same spectral region, which could lead to a poor
quality determination of the component area. Also, the baseline
correction may produce larger errors when the integrated area
method is used. In order to determine the concentration of
EFAD from the peak intensity, we have first determined the
molar extinction coefficient for the relevant band (see Table 1)
by recording the infrared spectra of EFAD solutions in CCl
with known concentrations. The use of an apolar solvent such
as CCl is not expected to have a significant influence on the
molar extinction coefficient.
4
Fig. 2 IR spectra of IEFAD in the CO
2
-rich phase at 5 MPa (black),
4
◦
1
5 MPa (red) and 20 MPa (blue), at 70 C.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 2205–2213 | 2207

View Article Online
Table 1 Molar extinction coefficient of IEFAD, TEFAD and CO
2
for different absorption bands
TEFAD
IEFAD
CO
2
Group frequency
Wave number/cm
n
C
O
n
C–H
n
C–H+ dC–H
2nC–H
n
C
O
n
C–H
n
C–H+ dC–H
2nC–H
n
1+ 2n2+
n
3
-1
1736
1069
2857
821
4263
4.73
5690
2.60
1736
1127
2863
829
4263
7.42
5711
1.51
4960
0.21
-1
-1
e (L mol . cm )
In order to estimate the EFAD concentration in the EFAD-
rich phase swelled with CO , the same procedure as described
above was followed. In this case, the peak height of the band
centred at 5700 cm which is associated to an overtone of the
CH vibrations of EFAD was considered. The molar extinction
coefficient for this band (reported in Table 1) was determined
from the spectrum measured under ambient conditions for pure
EFAD for which the density (or concentration) was known.
the density has been studied for different temperatures (Fig.
5). Whatever the temperature, the higher the CO density, the
higher the solubility of both EFADs. For a given density, the
temperature effect seems to be almost negligible. Therefore,
it appears that the driving force that affects the solubility of
2
2
-
1
such EFAD molecules in CO is the density of the supercritical
2
medium. In order to examine this point, we have compared
on Fig. 6 the solubility of TEFAD and IEFAD with that of a
glyceride series.
Finally, the concentration of CO
2
(CCO2) incorporated into
the EFAD phase was calculated using the peak height of the
-
1
n
1
+2n
The molar absorption coefficient displayed in Table 1 was
estimated by measuring the infrared spectra of supercritical CO
for a range of temperatures and pressures, and the corresponding
2
+n
3
band of CO
2
centred at 4950 cm .
2
3
5
densities (or concentrations) were available from the literature.
It was assumed that the molar absorption coefficient of this peak
was the same when CO swells the EFAD molecule.
2
Solubility of epoxy fatty acid diester (EFAD) in CO
2
. Fig. 4
displays the concentration of TEFAD and IEFAD in the CO
2
-
rich phase as a function of pressure at different temperatures.
Whatever the temperature, the solubility of both EFADs is
negligible at pressures below 10 MPa and then increases strongly
above this threshold. This effect is particularly clear at low
temperature. Although the concentration values are quite low,
the EFADs studied have a significant solubility in supercritical
Fig. 5 Concentration of TEFAD and IEFAD in the CO -phase as a
function of the CO density.
2
2
CO
2
. For a given pressure, a temperature increase causes a
decrease in solubility of both EFADs. For a given (P, T) couple,
the solubility of TEFAD in the CO
that of IEFAD.
2
-rich phase is higher than
◦
2
Fig. 6 Ln[EFAD] as a function of the density of CO at 50 C.
◦
36
Comparison with literature data (T = 50 C) .
This series is composed of oleic acid, monolein (ester of
glycerol with one C18:1 chain), diolein (ester of glycerol with two
C18:1 chains) and triolein (triglyceride with three C18:1 chains).
The latter data was reported in the literature as a function of the
Fig. 4 Concentration of TEFAD and IEFAD in the CO
function of the pressure, at different temperatures (70 C and 80
2
-phase as a
◦
◦
C
for IEFAD). Insert: concentration of TEFAD in the CO
function of the pressure and temperature.
2
-phase as a
◦
36
density of supercritical CO at T = 50 C. Concerning the
2
In order to discriminate the temperature and pressure effects,
the concentration of TEFAD and IEFAD as a function of
glyceride series, the difference in polarity and molecular weight
of the tested molecules was shown to affect the vapor pressures
2
208 | Green Chem., 2010, 12, 2205–2213
This journal is © The Royal Society of Chemistry 2010

View Article Online
and thus the solubility in CO
2
. Monoglyceride is logically more
soluble in CO than diglyceride and triglyceride because of its
2
lower carbon number. The solubilities of TEFAD and IEFAD
were found to be in-between oleic acid and monoglyceride. The
-
1
lower molecular weight of TEFAD (468 g mol ) versus that of
-
1
IEFAD (664 g mol ) explains its higher solubility in CO
2
. Due
to its lower molecular weight (282 g mol ), oleic acid is the most
volatile and thus the most soluble in CO of the entire series.
-
1
2
-
1
The lower solubility of monoglyceride (355 g mol ) compared
to oleic acid is also explained by its higher polarity due to the
presence of the two hydroxyl groups. Despite its lower molecular
weight, its solubility is close to that of TEFAD and IEFAD.
Scheme 2 Reaction scheme of IEFAD and TEFAD carbonation.
Swelling of epoxy fatty acid diester by CO
2
. The swelling
of TEFAD and IEFAD in the EFAD-rich phase as a function
◦
◦
of CO pressure at T = 50 C and 80 C for IEFAD and at
2
◦
◦
T = 50 C to 120 C for TEFAD is reported in Fig. 7. The
swelling S is the ratio of the initial concentration (C ) versus
the concentration (C) at a given CO pressure: [S = (C /C) -
]. The highest swelling is obtained for TEFAD in comparison
0
2
0
1
to IEFAD. The swelling of both EFADs increases significantly
with pressure up to a value of about 60% reached for a pressure
◦
of 20 MPa at 50 C. For a given pressure, an increase of the
temperature yields a decrease of the EFAD swelling. This data
clearly shows that CO
2
is able to penetrate into the epoxidised
oils (EFADs), resulting in a swelling of the EFAD phase which
37
is significantly higher than that usually measured for polymers
but lower than that reported for common solvents.
38
Fig. 8 In situ IR monitoring of the carbonation reaction of IEFAD
◦
into ICFAD at 80 C and 10 MPa.
Fig. 7 Swelling of TEFAD and IEFAD as a function of the CO
2
pressure at different temperatures.
Kinetic study of epoxy fatty acid diester carbonation reaction
◦
Fig. 9 IR spectra of in situ carbonation monitoring of IEFAD at 80 C
The kinetics of the carbonation reaction (Scheme 2) was studied
and 10 MPa. The spectrum of the IEFAD precursor (black) and the
spectrum recorded after 25 min reaction time (red). Insert: changes
in the peak height of the carbonate band (772 cm ) and epoxy band
by varying the pressure of CO
and the temperature (from 60 C up to 140 C). Fig. 8 and Fig. 9
2
(from 5 MPa up to 18.5 MPa)
◦
◦
-1
illustrate the spectral changes of EFAD in the wavenumber range
-1
(
826 cm ) at different reaction times.
1
6
00–4500 cm- that occur with the carbonation of epoxy groups.
One can detect the formation of two bands corresponding to
few minutes into the reaction). In order to determine the EFAD
conversion into the carbonated fatty acid diester (CFAD), the
height of the peak at 772 cm was converted into a percentage.
-
1
-1
carbonate groups, one at 1803 cm and one at 772 cm . Another
-
1
significant peak at 826 cm is associated to the disappearance of
the epoxy groups. The reaction progress was monitored by the
-1
-
1
height of the carbonate band at 772 cm and the epoxy band
Influence of the temperature. First, the pressure was fixed
at 10 MPa to study the influence of the temperature on the
at 826 cm- (the peak at 1803 cm becomes saturated after a
1
-1
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 2205–2213 | 2209

View Article Online
kinetics of carbonation. All the runs were stopped after 100 min
into the reaction by depressurizing the cell. Fig. 10 (blue tags)
displays the conversion of EFAD into CFAD—based on the
isophoronediamine (IPDA) (stoichiometric ratio, [carbonate] =
[NH ]), at different temperatures, in the absence of catalyst
2
(Scheme 3). As it is illustrated in Fig. 11 for the case of ICFAD
and IPDA as monomers, the polyadditions were monitored by
in situ ATR-FTIR spectroscopy, using the disappearance of the
-
1
intensity of the growing carbonate band at 772 cm —for both
TEFAD and IEFAD at different temperatures ranging from
◦
◦
-1
6
0 C to 140 C. As expected, the temperature has a dramatic
carbonate bands at 1803 and 772 cm and the appearance of
-
1
effect on the reaction kinetics for both EFADs. Fig. 10 clearly
shows the faster kinetics of TEFAD conversion with respect
to IEFAD, a phenomenon explained by the higher reactivity
of the terminal epoxide groups compared to internal ones. As
new bands at 3330, 1714 and 1530 cm , assigned to the hydroxyl
group, carbonyl group and N–H deformation of the urethane
function, respectively.
an example, after 100 min, 92% of the TEFAD is converted at
◦
1
00 C in comparison to 45% for IEFAD. As a conclusion, a
◦
temperature increase of up to 140 C is required to get complete
conversion with IEFAD, compared to 120 C in the case of
◦
TEFAD.
Scheme 3 Polyurethane synthesis by polyaddition between CFAD and
diamines.
Fig. 10 Percentage of carbonation of TEFAD and IEFAD respectively
◦
as a function of the pressure at 120 C (in red) and as a function of the
temperature at 10 MPa (in blue), after 100 min into the reaction.
Influence of the pressure. Second, the temperature was fixed
◦
at 120 C and the carbonation was performed at pressures
ranging from 5 MPa to 18.5 MPa. The conversion of IEFAD
◦
Fig. 11 In situ ATR-IR monitoring of the polyaddition at 70
between ICFAD and IPDA as a function of time.
C
into ICFAD is strongly affected by the CO
2
pressure (see Fig.
1
0, red tags). The conversion is maximal at 10 MPa (90%) and
Fig. 12 represents the decrease of the carbonate band ab-
above this optimal value, increasing the pressure does not favor
the conversion of epoxide groups into carbonate groups.
In the case of TEFAD, the conversion of epoxy groups into
-
1
◦
◦
sorbance at 1803 cm as a function of time, at 70 C and 110 C
in the case of CFAD/IPDA and CFAD/EDA polyadditions. As
expected, an increase of the temperature eases the conversion
of the carbonate groups into urethane functions. Fig. 12 also
highlights the higher reactivity of TCFAD in comparison to
ICFAD proving that terminal carbonate groups are much more
reactive than internal ones. In the case of the ICFAD/IPDA
carbonate ones is almost unaffected by the CO pressure at
2
this temperature and varies between 94% and 100%. As already
discussed, this is explained by the higher reactivity of the epoxy
groups in terminal positions. Whatever the conditions (P, T,
1
00 min), the conversion of TEFAD into TCFAD is always
◦
system, the conversion is complete after 12 h at 110 C in
higher than that of IEFAD into ICFAD.
comparison to 9 h with TCFAD/IPDA. One can also notice
that the rate of carbonate conversion is much more sensitive to
the temperature increase in the case of ICFAD polymerizations.
Finally, when EDA was used as comonomer, the conversion of
the carbonate function did not reach completion (pink curve),
Synthesis of polyurethanes from CFAD and diamines
Kinetics of CFAD polymerization. ICFAD and TCFAD were
both polymerized in bulk with ethylenediamine (EDA) and
2
210 | Green Chem., 2010, 12, 2205–2213
This journal is © The Royal Society of Chemistry 2010

View Article Online
Fig. 12 Kinetics of CFAD/IPDA and CFAD/EDA polyadditions as
◦
◦
a function of time at 70 C and 110 C determined by the decrease of
-1
the carbonate band at 1803 cm .
and this behavior is explained by the occurrence of side reactions
(
see next section).
Influence of the diamine structure. In recent work dealing
Fig. 13 ATR-IR spectra of ICFAD (blue), polyurethane formed by
with PU synthesis through a carbonate route, Petrovic and co-
workers reported the occurrence of a side reaction between the
diamines used (EDA and butylenediamine (BDA)) with the
ester function of the vegetable oil, resulting in the formation
◦
reaction of ICFAD with IPDA at 70 C (red) and polyurethane formed
◦
by reaction of ICFAD with EDA at 70 C (black).
39
of amide groups and thus affecting the final polymer structure.
This reaction was favoured by increasing the temperature of
◦
◦
polymerization (from 70 C to 100 C). In this work, the
formation of amide functions was found to be affected by the
type of diamine used. Indeed, while no amide groups were
detected in the case of IPDA, in contrast, one could observe that
some of the ester groups were converted into amide functions
when EDA was used as the comonomer. The formation of
these amide groups was revealed by IR spectroscopy with the
-
1
appearance of a specific band at 1655 cm as indicated in Fig.
3.
The H NMR spectrum of ICFAD together with that of
1
1
the PU obtained from reaction between ICFAD and EDA are
represented in Fig. 14. New peaks at 3.6 and 4.6 ppm reveal
the formation of the urethane linkage (e) and hydroxyl group
(
f). Nevertheless, signals at 4.2 and 4.5 ppm corresponding
to the protons a of the carbonate group have not completely
disappeared at the end of the reaction, which confirms the
presence of residual carbonate groups. The presence of amide
groups is characterised by the appearance of peaks at 2.1 ppm,
related to protons g. Integration of these peaks reveals that 22%
of amide groups are formed. In conclusion, IPDA appeared to
be advantageously selective for carbonate versus ester groups,
avoiding the formation of amide linkages. Such a side reaction is
favored when using more reactive primary amines, such as EDA,
in agreement with literature data.
1
Fig. 14 H NMR spectra of ICFAD and of PU obtained after reaction
◦
PU average molecular weights. The so-formed PUs were
found to be soluble in THF enabling the determination of their
average molecular weights in this solvent, against a polystyrene
calibration. The data are collected in Table 2. For a given
between ICFAD and EDA at 70 C.
in comparison to ICFAD, the highest PU molecular weights
were obtained with the TCFAD/IPDA monomer couple at
110 C (M
◦
◦
-1
couple of monomers, raising the temperature from 70 C to
w
= 13 500 g mol ). The dispersity M
w
/M of the
n
◦
1
10 C leads to an increase in polyurethane average molecular
polyurethane samples remain narrow, below 1.5, that can be
weight. In agreement with the higher reactivity of TCFAD
probably explained by incomplete conversion. It is worth noting
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 2205–2213 | 2211

View Article Online
Table 2 Characteristics of TCFAD- and ICFAD-based polyurethanes
MPa). TEFAD was found to be more soluble than IEFAD in the
CO
2
-rich phase, certainly because of its lower molecular weight
was also slightly
Total
conv. mol
w
M /g
-1b
◦
a
c
◦
g
T / C
d
and greater volatility. The concentration of CO
2
Monomers T/ C
M
w
/M
n
% Amide
higher in the TEFAD-rich phase than in the IEFAD-rich one.
This behavior could partially govern the kinetics of carbonation,
performed in the presence of TBABr (3 wt%) as catalyst. Indeed,
for a selected (P; T), TEFAD was more easily converted into
TCFAD compared to IEFAD. The polyurethanes obtained by
polyaddition of CFAD monomers with diamines (IPDA and
ICFAD/EDA 70
ICFAD/IPDA 70
ICFAD/IPDA 110 12 h
TCFAD/IPDA 70 12 h
TCFAD/IPDA 110 9 h
3 h
20 h
4 300
9 100
11 700
10 700
13 500
1.1
1.3
1.4
1.4
1.5
22
—
—
—
—
-25
-21
-19
-16
-13
a
-1
At complete disappearance of the carbonate band at 772 cm , except
-
1
b
EDA) exhibit molecular weights up to 13 500 g mol and glass
for poly(ICFAD/EDA) because of residual carbonate groups. SEC,
solvent THF, PS calibration. Determined by H NMR (integration of
◦
c
1
transition temperatures around -15 C. Interestingly, the side
d
◦
◦
peaks at 2.1 ppm). DSC, temperature ramp from -100 C to 200 C,
reaction between amine and ester functions leading to amide
linkages was not observed when the secondary diamine IPDA
was used as the comonomer.
◦
-1
at 10 C min .
that PUs obtained from polyadditions of ICFAD with EDA
exhibit much lower molecular weights, explained by a loss
of amine functions through the formation of amide linkages.
Fig. 15 shows the SEC traces of polyurethanes obtained from
TCFAD and IPDA (poly[TCFAD-IPDA]) and from ICFAD
References
1
K. C. Frisch, J. Polym. Sci., Part A: Polym. Chem., 1991, 29, 1834–
835.
1
2
F. Seniha G u¨ ner, Y. YagcI and A. Tuncer Erciyes, Prog. Polym. Sci.,
2006, 31, 633–670.
◦
3
4
5
6
A. Gandini, Macromolecules, 2008, 41, 9491–9504.
Z. S. Petrovi c´ , Polym. Rev., 2008, 48, 109–155.
V. Sharma and P. P. Kundu, Prog. Polym. Sci., 2008, 33, 1199–1215.
Z. S. Petrovic, A. Guo and W. Zhang, J. Polym. Sci., Part A: Polym.
Chem., 2000, 38, 4062–4069.
and IPDA (poly[ICFAD-IPDA]), both at 110 C. It also reveals
the absence of residual monomer in the PU samples.
7
8
9
Y. H. Hu, Y. Gao, D. N. Wang, C. P. Hu, S. Zu, L. Vanoverloop and
D. Randall, J. Appl. Polym. Sci., 2002, 84, 591–597.
L. Monteavaro, E. da Silva, A. Costa, D. Samios, A. Gerbase and C.
Petzhold, J. Am. Oil Chem. Soc., 2005, 82, 365–371.
Z. Lozada, G. J. Suppes, Y. C. Tu and F. H. Hsieh, J. Appl. Polym.
Sci., 2009, 113, 2552–2560.
1
1
1
0 A. Guo, D. Demydov, W. Zhang and Z. S. Petrovic, J. Polym. Environ.,
002, 10, 49–52.
1 Z. S. Petrovic, A. Guo, I. Javni, I. Cvetkovi and D. P. Hong, Polym.
Int., 2008, 57, 275–281.
2 P. Tran, D. Graiver and R. Narayan, J. Am. Oil Chem. Soc., 2005,
82, 653–659.
2
Fig. 15 SEC traces of TCFAD (red), ICFAD (black), PU obtained by
◦
polyaddition of ICFAD with IPDA at 110 C (green) and PU obtained
13 M. Ionescu, Z. S. Petrovi c´ and X. Wan, J. Polym. Environ., 2007, 15,
◦
237–243.
by polyaddition of TCFAD with IPDA at 110 C (blue). Solvent: THF;
1
1
1
4 S. Gryglewicz, W. Piechocki and G. Gryglewicz, Bioresour. Technol.,
003, 87, 35–39.
5 L. Hojabri, X. Kong and S. S. Narine, Biomacromolecules, 2009, 10,
84–891.
PS calibration.
2
8
6 D. V. Palaskar, A. Boyer, E. Cloutet, C. Alfos and H. Cramail,
Biomacromolecules, 2010, 11, 1202–1211.
Thermo-mechanical analysis of PU samples. The linear PUs
were analyzed by differential scanning calorimetry (DSC). All
17 US Pat., 6 120 905, 2006.
8 US pat., 6 120 905, 2000.
1
samples exhibit one glass transition temperature, T
g
, ranging
from -25 C to -13 C. PUs obtained from ICFAD exhibit a
that is systematically lower than that of PUs from TCFAD,
a phenomenon explained by the slightly lower molecular weight
◦
◦
19 H. Tomita, F. Sanda and T. Endo, J. Polym. Sci., Part A: Polym.
Chem., 2001, 39, 162–168.
T
g
2
0 G. Rokicki and A. Piotrowska, Polymer, 2002, 43, 2927–2935.
21 B. Tamami, S. Sohn and G. L. Wilkes, J. Appl. Polym. Sci., 2004, 92,
83–891.
2 K. M. Doll and S. Z. Erhan, J. Agric. Food Chem., 2005, 53, 9608–
614.
8
of these PUs but also by the presence of pendant chains (C
which can act as plasticisers.
8
17
H )
2
9
As expected, the T
g
of all these linear PUs are lower than
23 K. M. Doll and S. Z. Erhan, Green Chem., 2005, 7, 849–854.
24 P. G. Parzuchowski, M. Jurczyk-Kowalska, J. Ryszkowska and G.
Rokicki, J. Appl. Polym. Sci., 2006, 102, 2904–2914.
those of cross-linked soy-based PUs obtained from carbonated
◦
39
triglycerides and EDA, that display a T around 20 C.
g
2
5 B. M. C. Soares, F. M. C. Gamarra, L. C. Paviani, L. A. G. Gon c¸ alves
and F. A. Cabral, J. Supercrit. Fluids, 2007, 43, 25–31.
2
2
6 G. Madras, Fluid Phase Equilib., 2004, 220, 167–169.
Conclusions
7 G. Madras, C. Kulkarni and J. Modak, Fluid Phase Equilib., 2003,
2
09, 207–213.
In this study, novel linear polyurethanes were synthesized by
bulk polyaddition of diamines with vegetable-based biscarbon-
ates, namely ICFAD and TCFAD, produced from oleic acid
methyl ester. ICFAD and TCFAD were obtained by the reaction
2
8 V. Vasconcellos and F. Cabral, J. Am. Oil Chem. Soc., 2001, 78, 827–
8
29.
29 H. Sovov a´ , M. Zarev u´ cka, M. Vacek and K. Str a´ nsk y´ , J. Supercrit.
Fluids, 2001, 20, 15–28.
3
3
0 J.-N. Jaubert, P. Borg, L. Coniglio and D. Barth, J. Supercrit. Fluids,
001, 20, 145–155.
1 U. Guclu and F. Temelli, Ind. Eng. Chem. Res., 2000, 39, 4756–4766.
of their epoxide precursors (IEFAD and TEFAD) with CO
phase behavior of the EFAD/CO mixtures was first studied
with respect to temperature (70–110 C) and pressure (5–20
2
. The
2
2
◦
32 Fr. Pat., 09 58219, 2010.
2
212 | Green Chem., 2010, 12, 2205–2213
This journal is © The Royal Society of Chemistry 2010

View Article Online
3
3
3
3 V. Martinez, S. Mecking, T. Tassaing, M. Besnard, S. Moisan,
36 O¨ . G u¨ c¸ l u¨ - U¨ st u¨ ndag and F. Temelli, J. Supercrit. Fluids, 2006, 38,
F. Cansell and C. Aymonier, Macromolecules, 2006, 39, 3978–
275–288.
3
979.
37 D. L. Tomasko, H. Li, D. Liu, X. Han, M. J. Wingert, L. J. Lee and
K. W. Koelling, Ind. Eng. Chem. Res., 2003, 42, 6431–6456.
38 A. Kordikowski, A. P. Schenk, R. M. Van Nielen and C. J. Peters, J.
Supercrit. Fluids, 1995, 8, 205–216.
39 I. Javni, P. H. Doo and Z. S. Petrovic, J. Appl. Polym. Sci., 2008, 108,
3867–3875.
4 G. Herzberg, in Molecular Spectra and Molecular Structure: Infrared
and Raman Spectra of Polyatomic Molecules, D. Van Nostrand
Company, Inc., Princeton, New Jersey, 1956, pp. 273–277.
5 National Institute of Standards and Technology, http://webbook.
nist.gov/chemistry/.
This journal is © The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 2205–2213 | 2213