Model studies and the ADMET polymerization of soybean oil

Article abstract of DOI:10.1007/s11746-002-0509-3

Grubbs' ruthenium catalyst 2 has been employed in model studies of the acyclic diene metathesis (ADMET) polymerization of soybean oil. In the presence of 0.1 mol% of catalyst 2, the ADMET polymerization of ethylene glycol dioleate afforded the isomerized (E)-dioleate (27%), dimer (18%), trimer (13%), tetramer (7%), pentamer (5%), hexamer (4%), heptamer (4%), and 9-octadecene (21%). Only a trace of any intramolecular cyclized product was formed. Under the same conditions, glyceryl trioleate underwent ADMET polymerization to produce dimer, trimer, tetramer, pentamer, and monocyclic oligomers, with monocyclic oligomers predominating. The high number of repeat units in the monocyclic oligomers (n ? 6, 10, and 21) indicates that cross-linking occurs readily in this process. Based on our model system studies, we have examined the ADMET polymerization of soybean oil and succeeded in producing polymeric materials ranging from sticky oils to rubbers.

Full text of DOI:10.1007/s11746-002-0509-3

Model Studies and the ADMET Polymerization of Soybean Oil
Qingping Tian and Richard C. Larock*
Department of Chemistry, Iowa State University, Ames, Iowa 50011-3111
ABSTRACT: Grubbs’ ruthenium catalyst 2 has been employed catalyst, Grubbs’ (Cy3P)2Cl2Ru=CHPh catalyst, for this
in model studies of the acyclic diene metathesis (ADMET) poly-
merization of soybean oil. In the presence of 0.1 mol% of cata-
lyst 2, the ADMET polymerization of ethylene glycol dioleate
afforded the isomerized (E)-dioleate (27%), dimer (18%), trimer
(13%), tetramer (7%), pentamer (5%), hexamer (4%), heptamer
(4%), and 9-octadecene (21%). Only a trace of any intramolec-
ular cyclized product was formed. Under the same conditions,
glyceryl trioleate underwent ADMET polymerization to produce
dimer, trimer, tetramer, pentamer, and monocyclic oligomers,
with monocyclic oligomers predominating. The high number of
repeat units in the monocyclic oligomers (n ≅ 6, 10, and 21) in-
process (10). This approach proved highly efficient for the
olefin metathesis of soybean and many other natural oils and
provided much more thoroughly characterized metathesized
soybean drying oils. At the time, the exact nature of the
metathesis process was not well understood. We now wish to
report model studies on the metathesis and co-metathesis of
soybean oil that fully explain the metathesis process and that
have led to higher-M.W. polymeric materials with consider-
able industrial potential. This work nicely complements the
earlier work of Finkel’shtein et al. (11), who have also exam-
dicates that cross-linking occurs readily in this process. Based ined the co-metathesis of synthetic and natural oils using the
on our model system studies, we have examined the ADMET
polymerization of soybean oil and succeeded in producing
polymeric materials ranging from sticky oils to rubbers.
Paper no. J9759 in JAOCS 79, 479–488 (May 2002).
WCl6/Me4Sn catalyst.
Olefin metathesis has also been employed in the synthesis
of polymers, primarily through ring-opening metathesis poly-
merization and acyclic diene metathesis polymerization
(ADMET). The latter process involving dienes allows high-
M.W. polymeric products to be produced by selectively remov-
ing a volatile by-product, such as ethylene, during the reaction
(Eq. 2) (12–14). The choice of catalyst has proven very impor-
tant in ADMET polymerization. Thus, Schrock’s well-defined
KEY WORDS: ADMET polymerization, Grubbs’ catalyst,
olefin metathesis, soybean oil.
Olefin metathesis is a reaction in which olefins are formally tungsten and molybdenum alkylidenes, such as 1 (15–17), and
cleaved at the carbon-carbon double bond and new olefins re- more recently Grubbs’ ruthenium alkylidene complex, 2
sult by recombination of the fragments (Eq. 1) (1,2). A wide (18–20), have proven highly successful in ADMET polymer-
variety of transition metal complexes have been shown to cat- ization. These Lewis acid-free catalysts show higher activities
alyze this process. Particularly important are those involving in ADMET polymerization than the classical metathesis cata-
W, Mo, Re, and Ru. Until recently, the most widely used cata- lysts, and the reaction can be carried out in short reaction times
lyst system employed WCl6 and an alkylmetal or a Lewis acid. at room temperature while accommodating a variety of func-
tional groups (21) (Schemes 1 and 2).
catalyst
2 CH3CH=CHC2H5
[1]
CH3CH=CHCH3 + C2H5CH=CHC2H5
catalyst
H C
2
CH2
=CH–R–CH= n
n H2C=CH–R–CH=CH2
[2]
The olefin metathesis of esters of unsaturated FA has been ex-
amined. Boelhouwer and co-workers have reported the olefin
metathesis of methyl oleate (3) and the more highly unsatu-
+ (n − 1)H2C=CH2
We envisioned that the ADMET process could be em-
rated esters of linoleic acid and linolenic acid (4–6) using the ployed in the polymerization of soybean oil. Since the
catalyst WCl6/Me4Sn (3). Using the same catalyst system, ADMET process proceeds as a step polymerization, the
Kohashi and Foglia (7) have examined the co-metathesis of ADMET polymerization of soybean oil should be better con-
methyl oleate with other unsaturated diesters. Schuchart in- trolled than the previously used thermal polymerization
vestigated the co-metathesis of methyl oleate and ethylene in process (22,23). It was our hope that this process would con-
the presence of Re3O7/SiO2·Al2O3/B2O3 (8).
vert soybean oil to useful new biodegradable materials. We
The WCl6/Me4Sn-catalyzed metathesis of soybean oil, now report the success of that effort.
which consists mostly of the TG of oleic (~23%), linoleic
(~51%), and linolenic acids (~7%) has been reported to pro-
duce an improved drying oil (9). However, this inefficient and
EXPERIMENTAL PROCEDURES
environmentally hazardous procedure, and lack of any real General. All 1H and 13C NMR spectra were recorded at 300
physical properties for the resulting oil, encouraged us re- and 75.5 MHz, respectively. Gel permeation chromatography
cently to look at the use of a more modern olefin metathesis (GPC) analyses were carried out with the use of a Waters gel
permeation system (410 refractive index detector; Milford,
*To whom correspondence should be addressed. E-mail larock@iastate.edu
Copyright © 2002 by AOCS Press
479
JAOCS, Vol. 79, no. 5 (2002)

480
Q. TIAN AND R.C. LAROCK
oleic acid (10 g, 35.4 mmol) in 50 mL of CCl4 at 0°C. This
was followed by the addition of 4-(dimethylamino)pyridine
(4.32 g, 35.4 mmol). Most of these components dissolved
after stirring for 30 min. A solution of N,N-dicyclohexylcar-
bodiimide (7.31 g, 35.4 mmol) in dry CCl4 was added to the
mixture, which was then allowed to stir at room temperature
for 5 h. The resulting mixture was filtered, and the precipitate
was washed with CCl4. Evaporation of the filtrate under re-
duced pressure gave the crude product, which was purified by
flash chromatography (20:1 hexanes/EtOAc) to afford a col-
orless liquid (7.13 g, 12.1 mmol, 68%): 1H NMR (CDCl3) δ
0.87 (t, J = 6.2 Hz, 6H), 1.27 (m, 40H), 1.61 (m, 4H), 2.00 (m,
8H), 2.31 (t, J = 7.5 Hz, 4H), 4.26 (s, 4H), 5.24–5.36 (m, 4H);
13C NMR (CDCl3) δ 14.1, 22.7, 24.9, 27.1, 27.2, 29.1, 29.13,
29.2, 29.3, 29.6, 29.7, 29.8, 31.9, 34.1, 62.0, 129.7, 130.0,
173.5 (one sp3 carbon missing due to overlap); IR (CDCl3)
3002, 2952, 1742, 1456 cm−1; MS (APCI) m/z 591 (MH)+.
Glyceryl trioleate. Glyceryl trioleate was prepared accord-
ing to a literature procedure (24). Glycerol (0.429 g, 4.66
mmol) was added to a solution of oleic acid (5.0 g, 17.7
mmol) in 25 mL of CCl4 at 0°C. This was followed by the ad-
dition of 4-(dimethylamino)pyridine (2.162 g, 17.7 mmol).
Most of these components dissolved after stirring for 30 min.
A solution of N,N-dicyclohexylcarbodiimide (3.652 g, 17.7
mmol) in 30 mL of dry CCl4 was added to the mixture, which
was then allowed to stir at room temperature for 6 h. The re-
sulting mixture was filtered. and the precipitate was washed
with CCl4. Evaporation of the filtrate under reduced pressure
gave the crude product, which was purified by flash chroma-
tography (20:1 hexanes/EtOAc) to afford a colorless liquid
(2.10 g, 2.37 mmol, 51%): 1H NMR (CDCl3) δ 0.87 (t, J =
6.6 Hz, 9H), 1.28 (m, 60H), 1.60 (m, 6H), 1.99 (m, 12H), 2.30
(dt, J = 1.5, 6.0 Hz, 6H), 4.13 (dd, J = 6.0, 12.0 Hz, 2H), 4.29
(dd, J = 6.0, 12 Hz, 2H), 5.24–5.39 (m, 7H); 13C NMR
(CDCl3) δ 14.1, 22.7, 24.8, 24.9, 27.1, 27.2, 29.0, 29.10,
29.13, 29.2, 29.3, 29.5, 29.7, 29.8, 31.9, 34.1, 34.2, 62.1,
68.9, 129.6, 129.7, 130.0, 172.8, 173.3 (14 sp3 carbons miss-
ing due to overlap); IR (CDCl3) 3002, 2924, 1744, 1463
cm−1; MS (APCI) m/z 886 (MH)+.
ADMET polymerization of ethylene glycol dioleate. In a ni-
trogen-filled dry box, catalyst 2 (2.0 mg, 2.43 µmol) was
weighed into a Schlenk tube with a magnetic stir bar. After
being capped with a stopcock, the flask was removed from the
dry box and attached to a manifold. Another Schlenk flask
loaded with ethylene glycol dioleate was also connected to the
manifold. The manifold was evacuated and filled with argon
three times. The two Schlenk flasks were then opened to the
manifold. Under a steady flow of argon, ethylene glycol di-
oleate (1.5 mL, 1.39 g, 2.36 mmol) was transferred to the flask
charged with catalyst 2. The flask was then switched to the vac-
uum line, and the reaction mixture was slowly warmed to 55°C
for 24 h while stirring. The flask was removed from the bath
and allowed to cool. Then CH2Cl2 (10 mL), ethyl vinyl ether
(0.1 mL), and BHT (15 mg) were added to the flask. Ethyl vinyl
ether and BHT were used to terminate the reaction and to sta-
bilize the polymer product, respectively. After 12 h of stirring,
N
RO
H
Mo=C
RO
Ph
R = CMe(CF3)2
1
SCHEME 1
Cl PCy3
H
Ru=C
Cl
Ph
PCy3
Cy = cyclohexyl
2
SCHEME 2
MA) coupled with a Wyatt miniDawn, a laser light-scattering
photometer (Wyatt Technology Corp., Santa Barbara, CA).
The chromatography system was equipped with three ultra-
styragel columns (Waters HR 1, 4, and 5). The M.W. were
calculated by calibration with polystyrene standards. THF
was utilized as the solvent, and the flow rate was 1.0 mL/min,
with the system equilibrated at 40°C.
GC–MS spectrometric analyses were performed using a
Finnigan Magnum Ion Trap Detector (San Jose, CA). The MS
system was configured in the electron impact ionization mode
with the automatic gain control feature turned on. A DB5-MS
column was used in the experiments. Atmospheric pressure
chemical ionization (APCI) MS experiments were performed
on a Finnigan TSQ 700 mass spectrometer equipped with a
Finnigan APCI ion source.
TLC was performed using commercially prepared 60-
mesh silica gel plates (Whatman K6F; Maidstone, United
Kingdom), and visualization was effected with short-wave-
length UV light (254 nm) or a basic KMnO4 solution [3 g
KMnO4 + 20 g K2CO3 + 5 mL NaOH (5%) + 300 mL H2O].
Reagents. All reagents were used directly as obtained com-
mercially unless otherwise noted. Ethylene glycol and
glycerol were purchased from Fisher Scientific (Pittsburgh,
PA). 1,9-Decadiene, ethyl vinyl ether, 2,6-di-tert-butyl-4-
methylphenol (BHT), N,N-dicyclohexylcarbodiimide, 4-(di-
methylamino)pyridine, and oleic acid were obtained from
Aldrich Chemical Co., Inc. (Milwaukee, WI). Catalysts 1 and
2 were provided by Professor Kenneth B. Wagener (Depart-
ment of Chemistry and Center for Macromolecular Science
and Engineering, University of Florida, Gainesville, FL). Cat-
alyst 2 was also purchased from Strem Chemicals, Inc. (New-
buryport, MA). 5-Hexenyl 4-pentenoate and di-5-pentenyl
1,4-benzenedicarboxylate were prepared according to previ-
ous literature procedures (16).
Ethylene glycol dioleate (5). Ethylene glycol dioleate (5)
was prepared according to a literature procedure (24). Ethyl-
ene glycol (1.097 g, 17.7 mmol) was added to a solution of
JAOCS, Vol. 79, no. 5 (2002)

THE ADMET POLYMERIZATION OF SOYBEAN OIL
481
A
3%
13%
22%
22%
14%
9%
A
39%
1%
B
BHT
B
Dimer
Trimer
Tetramer
Pentamer
Hexamer
Heptamer
42%
Light-brown
oil (48%)
White solid
(50%)
C
0.3%
14%
4%
Dimer
Trimer
8%
8%
A = CH3(CH2)7CH=CH(CH2)7CH3
B = E,E–CH3(CH2)7CH=CH(CH2)7CO2CH2CH2O2C(CH2)7CH=CH(CH2)7CH3
C = intramolecular cyclization product: (CH2)7
O
C
HC
CH
O
(CH2)7
C
O
O
Oligomer: (Z/E)–CH3(CH2)7CH=[CH(CH2)7CO2CH2CH2O2C(CH2)7CH=]nCH(CH2)7CH3
Dimer: n = 2; trimer: n = 3; tetramer: n = 4; pentamer: n = 5; hexamer: n = 6; heptamer: n = 7.
FIG. 1. Composition of the two fractions.
an additional 20 mL of CH2Cl2 was added to the solution, and 12H), 2.31 (t, J = 7.5 Hz, 8H), 4.26 (s, 8H), 5.24–5.42 (m, 6H);
the resulting solution was poured into 200 mL of rapidly stir- 13C NMR (CDCl3) δ 14.2, 22.7, 24.9, 27.2, 27.3, 29.0, 29.04,
ring methanol at 0°C. The white precipitate formed was then 29.1, 29.17, 29.2, 29.4, 29.5, 29.6, 29.7, 29.74, 29.8, 31.9,
separated from the solvents by centrifugation, followed by de- 32.6, 34.1, 62.0, 129.7, 129.9, 130.1, 130.2, 130.3, 130.5,
canting of the solvents, and then dried by pumping overnight. 173.6 (two sp3 carbons missing due to overlap); IR (CDCl3)
This procedure yielded a white solid (0.69 g, 50% weight re- 3052, 2952, 1739, 1461 cm−1; MS (APCI) m/z 930 (MH)+.
covery): m.p. 49–53°C. The solvent portions were collected,
(v) Trimer (Fig. 1). 1H NMR (CDCl3) δ 0.87 (t, J = 6.3 Hz,
concentrated and dried by pumping overnight, affording a 6H), 1.20–1.40 (m, 72H), 1.56–1.64 (m, 12H), 1.90–2.01 (m,
light-brown oil (0.66 g, 48% weight recovery). A portion of the 16H), 2.31 (t, J = 7.5 Hz, 12H), 4.26 (s, 12H), 5.24–5.42 (m,
white solid (0.2 g) was further partitioned by flash chromatog- 8H); 13C NMR (CDCl3) δ 14.1, 22.7, 24.9, 27.1, 27.2, 27.3,
raphy, yielding eight fractions (Fig. 1). The light-brown oil was 28.9, 29.0, 29.1, 29.15, 29.2, 29.3, 29.5, 29.55, 29.6, 29.72,
also processed by flash chromatography, affording six fractions 29.8, 31.9, 32.5, 32.6, 34.1, 62.0, 129.7, 129.8, 130.0, 130.2,
(Fig. 1). The following are the spectral data for these fractions: 130.3, 130.5, 173.6; IR (CDCl3) 3052, 2920, 1739, 1462
1
(i) (Z)- and (E)-9-octadecene (Fig. 1, Compound A). H cm−1; MS (APCI) m/z 1268 (MH)+.
NMR (CDCl3) δ 0.89 (t, J = 6.9 Hz, 6H), 1.20–1.33 (m, 24H),
(vi) Tetramer (Fig. 1). 1H NMR (CDCl3) δ 0.87 (t, J = 6.3
1.94–1.98 (m, 4H), 5.35–5.41 (m, 2H); 13C NMR (CDCl3) δ Hz, 6H), 1.20–1.40 (m, 88H), 1.56–1.64 (m, 16H), 1.90–2.10
14.2, 22.8, 27.7, 29.2, 29.4, 29.6, 29.7, 29.8, 32.0, 32.7, 129.9, (m, 20H), 2.31 (t, J = 7.5 Hz, 16H), 4.26 (s, 16H), 5.24–5.42
130.4; IR (CDCl3) 2952, 2849, 1463 cm−1; high-resolution (m, 10H); 13C NMR (CDCl3) δ 12.4, 14.1, 22.7, 24.9, 27.1,
MS (HRMS) m/z 252.2815 (calcd. for C18H36, 252.2817).
27.2, 27.3, 28.9, 29.0, 29.1, 29.16, 29.2, 29.3, 29.5, 29.55,
(ii) E,E-ethylene glycol dioleate (Fig. 1, Compound B). 1H 29.6, 29.72, 29.8, 31.9, 32.5, 32.6, 34.1, 62.0, 129.7, 129.8,
NMR (CDCl3) δ 0.87 (t, J = 6.3 Hz, 6H), 1.27 (m, 40H), 130.0, 130.2, 130.3, 130.5, 173.6; IR (CDCl3) 3052, 2920,
1.59–1.64 (m, 4H), 1.90–1.98 (m, 8H), 2.32 (t, J = 7.5 Hz, 4H), 1738, 1461 cm−1; MS (APCI) m/z 1605 (M)+.
4.26 (s, 4H), 5.24–5.36 (m, 4H); 13C NMR (CDCl3) δ 14.2,
(vii) Pentamer (Fig. 1). 1H NMR (CDCl3) δ 0.87 (t, J = 6.3
22.7, 24.9, 29.1, 29.12, 29.2, 29.23, 29.3, 29.5, 29.6, 29.7, 31.9, Hz, 6H), 1.20–1.40 (m, 104H), 1.56–1.64 (m, 20H),
32.6, 34.1, 62.0, 130.1, 130.5, 173.6 (one sp3 carbon missing 1.90–2.10 (m, 24H), 2.31 (t, J = 7.5 Hz, 20H), 4.26 (s, 20H),
due to overlap); IR (CDCl3) 3002, 2845, 1734, 1462 cm−1.
5.24–5.42 (m, 12H); 13C NMR (CDCl3) δ 14.2, 22.7, 24.9,
(iii) Compound C (Fig. 1). 1H NMR (CDCl3) δ 1.20–1.40 27.1, 27.2, 29.0, 29.1, 29.19, 29.2, 29.4, 29.5, 29.6, 29.7,
(m, 20H), 1.59–1.64 (m, 4H), 1.90–2.10 (m, 4H), 2.32 (t, J = 29.75, 29.8, 31.9, 32.6, 34.1, 62.0, 129.8, 129.9, 130.0, 130.2,
7.5 Hz, 4H), 4.30 (s, 4H), 5.30–5.40 (m, 2H); 13C NMR 130.3, 130.5, 173.6; IR (CDCl3) 3051, 2921, 1738, 1462
(CDCl3) δ 24.7, 27.8, 28.7, 28.8, 28.9, 31.9, 34.3, 62.0, 130.8, cm−1; MS (APCI) m/z 1946 (MH)+.
173.6; MS (APCI) m/z 339 (MH)+.
(viii) Hexamer (Fig. 1). 1H NMR (CDCl3) δ 0.87 (t, J = 6.3
(iv) Dimer (Fig. 1). 1H NMR (CDCl3) δ 0.87 (t, J = 6.3 Hz, Hz, 6H), 1.20–1.40 (m, 120H), 1.56–1.64 (m, 24H), 1.90–2.10
6H), 1.20–1.40 (m, 56H), 1.59–1.64 (m, 8H), 1.90–1.98 (m, (m, 28H), 2.31 (t, J = 7.5 Hz, 24H), 4.26 (s, 24H), 5.24–5.42
JAOCS, Vol. 79, no. 5 (2002)

482
Q. TIAN AND R.C. LAROCK
(m, 14H); 13C NMR (CDCl3) δ 14.2, 22.8, 24.9, 27.2, 29.1, 1.89 g, 2.14 mmol, assuming that the average M.W. of soy-
29.16, 29.2, 29.3, 29.39, 29.4, 29.6, 29.7, 29.8, 32.0, 32.6, bean oil is 884) was transferred to the flask charged with cat-
32.7, 34.2, 62.0, 129.9, 130.2, 130.3, 130.6, 173.6; IR (CDCl3) alyst 2. The flask was then switched to the vacuum line, and
3052, 2922, 1738, 1469 cm−1; MS (APCI) m/z 2282 (M)+.
the reaction mixture was slowly warmed to 55°C for the
(ix) Heptamer (Fig. 1). 1H NMR (CDCl3) δ 0.87 (t, J = 6.3 specified period of time while stirring. After that, the flask
Hz, 6H), 1.20–1.40 (m, 136H), 1.56–1.64 (m, 28H), 1.90–2.10 was removed from the bath and allowed to cool. If the result-
(m, 32H), 2.31 (t, J = 7.5 Hz, 28H), 4.26 (s, 28H), 5.24–5.42 ing polymer was a brown oil, procedure A was used to work
(m, 16H); 13C NMR (CDCl3) δ 14.2, 22.7, 24.9, 29.0, 29.1, up the reaction mixture. Procedure B was employed when the
29.18, 29.2, 29.3, 29.4, 29.51, 29.6, 29.71, 29.8, 31.9, 32.6, resulting product was a rubber.
32.7, 34.1, 62.0, 129.9, 130.2, 130.3, 130.5, 173.6; IR (CDCl3)
In procedure A, CH2Cl2 (20 mL), ethyl vinyl ether (0.2
mL), and BHT (30 mg) were added to the reactant flask. After
3052, 2921, 1737, 1461 cm−1; MS (APCI) m/z 2622 (MH)+.
ADMET polymerization of glyceryl trioleate or soybean 12 h of stirring, an additional 20 mL of CH2Cl2 was added to
oil. In a nitrogen-filled dry box, catalyst 2 (2.0 mg, 2.43 the solution, and the resulting solution was poured into 200
µmol) was weighed into a Schlenk tube containing a magnetic mL of rapidly stirring MeOH at 0°C. The stirring was contin-
stir bar. After being capped with a stopcock, the flask was re- ued until the product appeared free of color. Then the solvent
moved from the dry box and attached to a manifold. Another was decanted off, and the remaining solvent was evacuated.
Schlenk flask containing glyceryl trioleate or soybean oil was The residual material was then collected, dried by pumping
also connected to the manifold, and the mixture was degassed overnight, and recorded as MeOH/CH2Cl2-insoluble fraction
by three freeze–thaw cycles. The manifold was evacuated and B (Table 1). The MeOH solution was also concentrated, dried,
filled with argon three times. The two Schlenk flasks were and recorded as MeOH/CH2Cl2-soluble fraction A (Table 1).
then opened to the manifold. Under a steady flow of argon,
In procedure B, the resulting rubbery material was first
glyceryl trioleate (1.90 g, 2.14 mmol) or soybean oil (2 mL, processed utilizing procedure A. The soluble fraction was
TABLE 1
ADMET Polymerization of Soybean Oil
Isolated product (%)a
Oil Catalyst 2 Temp.
Time
(h)
Crude product
(%)
Fraction A Fraction B
Entry (mL) (mol%)
(°C)
Vac/Ar
Vac
(r)
(r)
Fraction C
1
2
3
4
5
6
7
8
1
1
1
1
1
2
2
2
1.5
1.4
1.4
1.4
1.4
1.5
0.2
0.2
r.t.
120
192
240
240
15
85
Brown rubber
r.t.
r.t.
r.t.
55
55
r.t.
r.t.
Vac
Vac
Ar
86
30
(0.36)
24
43
(1.23)
45
18
15
—
27
—
—
—
Brown, sticky rubber
87
Brown rubber
99
Dark-brown oil
86
Brown rubber
99
Deep-brown oil
89
(0.29)
40
(1.19)
58
(0.29)
23
(1.19)
36
Vac
Ar
(0.31)
37
(1.19)
66
24
(0.29)
17
(0.98)
71
Vac
Ar
240
240
Light-brown oil
100
(0.21)
38
(1.0)
62
Brown oil
(0.27)
(1.06)
9
10
11
12
13
14
15
2
2
2
2
2
2
2
0.2
0.2
0.1
0.1
0.1
0.01
0
55
55
55
55
55
55
55
Vac
Ar
63
62
82
Brown rubber
99
Brown oil
81
Brown rubber
90
Brown oil
98
Brown oil
100
Light-brown oil
99
18
(0.30)
43
(0.28)
24
(0.47)
23
(0.31)
41
(0.33)
100
(0.56)
99
30
(1.10)
55
(1.05)
31
(1.10)
67
(1.19)
57
(0.85)
—
34
—
26
—
—
—
—
Vac
Vac
Ar
192
24
24
Vac
Vac
168
212
—
Brown oil
(0.56)
aSee the Experimental Procedures section for the workup procedure. r.t., room temperature; Vac, vacuum.
JAOCS, Vol. 79, no. 5 (2002)

THE ADMET POLYMERIZATION OF SOYBEAN OIL
483
TABLE 2
Composition of the MeOH/CH2Cl2-Insoluble Fraction
Entry/
component
w/w
(%)
r
r a
(theory)
(exp)
n
Structural assignment
Comments
Yellow oil
White wax
White oil
White, sticky oil
White, sticky oil
White, very sticky oil
Very sticky oil
White, very sticky oil
1
2
3
4
5
6
7
8
5
10
6
3
3
15
17
31
0.39
0.70
0.85
0.99
1.08
1.17
1.28
1.43
0.44
0.66
0.80
0.89
0.95
1.00
1.04
1.06
1
2
3
4
5
6
7
8
Isomerized trioleate
Dimer
Trimer
Tetramer
Pentamer
Monocyclic oligomer
Monocyclic oligomer
Monocyclic oligomer
aThe theoretical value of r is calculated as 4n/(3n + 6), where n is the number of repeating units (see Eq. 4 in text).
recorded as fraction A (Table 1), and the residual rubber was 34.2, 62.1, 68.9, 129.8, 130.1, 130.2, 130.22, 130.3, 130.5,
further partitioned by Soxhlet extraction using CH2Cl2 as the 172.9, 173.3; IR (CDCl3) 3002, 2922, 1744, 1447 cm−1.
solvent. The CH2Cl2 solution was collected, dried, and
recorded as fraction B (Table 1). The remaining rubber was 0.80–0.90 (m, 21H), 1.20–1.42 (m, 204H), 1.50–1.65 (m,
recorded as fraction C (Table 1). 30H), 1.90–2.10 (m, 44H), 2.30 (dt, J = 1.5, 7.5 Hz, 30H),
(v) Component 5 (Table 2, entry 5). 1H NMR (CDCl3) δ
Fraction B, obtained from ADMET polymerization of 4.13 (dd, J = 6.0, 12.0 Hz, 10H), 4.29 (dd, J = 6.0, 12.0 Hz,
glyceryl trioleate, was further partitioned by flash chromatog- 10H), 5.24–5.39 (m, 27H); 13C NMR (CDCl3) δ 14.2, 22.8,
raphy and eight components were isolated. The following are 24.9, 27.2, 27.3, 29.0, 29.1, 29.15, 29.17, 29.2, 29.22, 29.26,
the spectral data for these components:
29.4, 29.5, 29.6, 29.66, 29.7, 29.72, 29.8, 31.9, 32.6, 32.7,
(i) Component 1 (Table 2, entry 1). 1H NMR (CDCl3) δ 0.87 34.1, 34.2, 62.1, 68.9, 129.8, 130.1, 130.2, 130.4, 130.6,
(m, 9H), 1.20–1.42 (m, 60H), 1.50–1.65 (m, 6H), 1.90–2.10 (m, 172.9, 173.3; IR (CDCl3) 3002, 2949, 1739 cm−1.
12H), 2.30 (dt, J = 1.5, 7.2 Hz, 6H), 4.13 (dd, J = 6.0, 12.0 Hz,
(vi) Component 6 (Table 2, entry 6). 1H NMR (CDCl3) δ
2H), 4.29 (dd, J = 6.0, 12.0 Hz, 2H), 5.24–5.39 (m, 7H); 0.80–0.90 (m, 3H), 1.20–1.42 (m, 36H), 1.50–1.65 (m, 6H),
13C NMR (CDCl3) δ 14.1, 22.7, 24.8, 24.9, 27.2, 29.0, 29.1, 1.90–2.10 (m, 8H), 2.30 (dt, J = 1.5, 7.5 Hz, 6H), 4.13 (dd, J
29.13, 29.2, 29.25, 29.3, 29.36, 29.5, 29.6, 29.63, 29.7, 29.75, = 6.0, 12.0 Hz, 2H), 4.29 (dd, J = 6.0, 12.0 Hz, 2H),
29.8, 31.9, 32.6, 32.7, 34.1, 34.2, 62.1, 68.9, 129.7, 130.1, 5.24–5.39 (m, 5H); 13C NMR (CDCl3) δ 14.1, 22.7, 24.8,
130.2, 130.22, 130.5, 172.9, 173.3; IR (CDCl3) 3002, 2921, 24.9, 27.2, 27.3, 29.0, 29.1, 29.14, 29.2, 29.23, 29.3, 29.5,
1744, 1461 cm−1; MS (APCI) m/z 886 (MH)+.
29.6, 29.67, 29.7, 29.75, 29.8, 31.9, 32.6, 32.7, 34.1, 34.2,
(ii) Component 2 (Table 2, entry 2). 1H NMR (CDCl3) δ 62.1, 68.9, 129.7, 129.8, 130.1, 130.2, 130.23, 130.5, 172.9,
0.80–0.90 (m, 12H), 1.20–1.42 (m, 96H), 1.50–1.65 (m, 12H), 173.3; IR (CDCl3) 2952, 1753, 1462 cm−1.
1.90–2.10 (m, 20H), 2.30 (dt, J = 1.5, 7.5 Hz, 12H), 4.13 (dd,
(vii) Component 7 (Table 2, entry 7). 1H NMR (CDCl3) δ
J = 6.0, 12.0 Hz, 4H), 4.29 (dd, J = 6.0, 12.0 Hz, 4H), 0.80–0.90 (m, 3H), 1.19–1.42 (m, 36H), 1.50–1.65 (m, 6H),
5.24–5.39 (m, 12H). 13C NMR (CDCl3) δ 14.1, 22.7, 24.8, 1.90–2.10 (m, 8H), 2.30 (dt, J = 1.5, 7.5 Hz, 6H), 4.13 (dd, J
24.9, 29.0, 29.1, 29.12, 29.16, 29.18, 29.23, 29.3, 29.4, 29.5, = 6.0, 12.0 Hz, 2H), 4.29 (dd, J = 6.0, 12.0 Hz, 2H),
29.6, 29.62, 29.7, 29.75, 29.8, 31.9, 32.6, 32.7, 34.1, 34.2, 5.22–5.39 (m, 5H); 13C NMR (CDCl3) δ 14.1, 22.7, 24.8,
62.1, 68.9, 129.6, 129.7, 129.8, 130.1, 130.2, 130.22, 130.5, 24.9, 27.2, 27.3, 29.0, 29.1, 29.15, 29.2, 29.23, 29.4, 29.5,
172.8, 173.3; IR (CDCl3) 3001, 2922, 1744 cm−1.
29.6, 29.63, 29.7, 29.76, 29.8, 31.9, 32.6, 32.7, 34.1, 34.2,
(iii) Component 3 (Table 2, entry 3). 1H NMR (CDCl3) δ 62.1, 68.9, 129.6, 130.1, 130.2, 130.22, 130.5, 172.9, 173.3;
0.80–0.90 (m, 15H), 1.20–1.42 (m, 132H), 1.50–1.65 (m, IR (CDCl3) 2946, 1738, 1456 cm−1.
18H), 1.90–2.10 (m, 28H), 2.30 (dt, J = 1.5, 7.5 Hz, 18H), 4.13
(viii) Component 8 (Table 2, entry 8). 1H NMR (CDCl3) δ
(dd, J = 6.0, 12.0 Hz, 6H), 4.29 (dd, J = 6.0, 12.0 Hz, 6H), 0.81–0.90 (m, 3H), 1.20–1.42 (m, 36H), 1.50–1.65 (m, 6H),
5.24–5.39 (m, 17H); 13C NMR (CDCl3) δ 14.1, 22.7, 24.8, 1.90–2.10 (m, 8H), 2.30 (dt, J = 1.5, 7.5 Hz, 6H), 4.13 (dd, J
24.9, 27.2, 27.3, 29.0, 29.1, 29.13, 29.2, 29.23, 29.3, 29.5, = 6.0, 12.0 Hz, 2H), 4.29 (dd, J = 6.0, 12.0 Hz, 2H),
29.6, 29.63, 29.66, 29.7, 29.75, 29.8, 31.9, 32.6, 32.7, 34.1, 5.22–5.38 (m, 5H); 13C NMR (CDCl3) δ 14.1, 22.7, 24.8,
34.2, 62.1, 68.9, 129.7, 129.8, 130.1, 130.2, 130.22, 130.5, 24.9, 27.2, 27.3, 29.0, 29.1, 29.15, 29.2, 29.3, 29.5, 29.6,
172.9, 173.3; IR (CDCl3) 3002, 2922, 1744, 1461 cm−1.
29.63, 29.7, 29.8, 31.9, 32.6, 32.7, 34.1, 34.2, 62.1, 68.9,
(iv) Component 4 (Table 2, entry 4). 1H NMR (CDCl3) δ 130.1, 130.2, 130.22, 130.5, 172.8, 173.3; IR (CDCl3) 2952,
0.80–0.90 (m, 18H), 1.20–1.42 (m, 168H), 1.50–1.65 (m, 1739, 1463 cm−1.
24H), 1.90–2.10 (m, 36H), 2.30 (dt, J = 1.5, 7.5 Hz, 24H), 4.13
(dd, J = 6.0, 12.0 Hz, 8H), 4.29 (dd, J = 6.0, 12.0 Hz, 8H),
RESULTS AND DISCUSSION
5.24–5.39 (m, 22H); 13C NMR (CDCl3) δ 14.1, 22.7, 24.8,
24.9, 27.2, 27.3, 29.0, 29.1, 29.12, 29.2, 29.24, 29.3, 29.5, Evaluation of ADMET polymerization catalysts. Since the
29.6, 29.63, 29.66, 29.7, 29.75, 29.8, 31.9, 32.6, 32.7, 34.1, catalyst is very important in ADMET polymerization, our ini-
JAOCS, Vol. 79, no. 5 (2002)

484
Q. TIAN AND R.C. LAROCK
CH3(CH2)7CH=CH(CH2)7CO2CH2
n
0.1 mol% of catalyst 2
CH3(CH2)7CH=CH(CH2)7CO2CH2 55°C, vacuum, 24 h
3
5
[3]
SCHEME 3
CH3(CH2)7CH =CH(CH2)7CO2CH2
+ (n − 1) CH3(CH2)7CH=CH(CH2)7CH3
9-octadecene
O
CH3(CH2)7CH =CH(CH2)7CO2CH2
n
After workup, two fractions, a light-brown oil (48%, w/w)
O
and a white, waxy solid (50%, w/w), were obtained. Each
fraction was then further partitioned by flash chromatogra-
phy. The composition of each fraction is shown in Figure 1.
The major components of the first fraction (a light-brown
oil) are compound A, 9-octadecene (39%, w/w), which is the
product expected to accompany formation of the polymer,
and compound B, the E,E-isomer of the starting dioleate
(42%, w/w). Only a trace amount of the intramolecular cy-
clization product, compound C (0.3%, w/w), was isolated
from this fraction. This indicates that the reaction does not
proceed to a significant extent by intramolecular cyclization,
but does follow the anticipated ADMET process to produce
oligomers. On the other hand, the second fraction (the white,
waxy solid) was mainly composed of the dimer (22%, w/w),
trimer (22%, w/w), and tetramer (14%, w/w). It is very inter-
esting that even pure pentamer, hexamer, and heptamer were
isolated from this fraction.
The overall product composition from the ADMET poly-
merization of the dioleate 5 was determined by combining the
isolated masses of each component from both fractions.
These combined masses were used to calculate an overall iso-
lated percent yield for each component. Thus, 27% of the
starting dioleate 5 was converted to the E,E-isomer B. The re-
maining dioleate (73%) underwent ADMET polymerization
to afford the following components: dimer (18%), trimer
(13%), tetramer (7%), pentamer (5%), hexamer (4%), hep-
tamer (4%), and 9-octadecene (21%).
All of these isolated compounds have been fully char-
acterized by 1H and 13C NMR spectroscopy, APCI MS, and
GPC. The 1H NMR spectra match the molecular structures
of the assigned compounds. For example, the dimer was
calculated to possess the protons CH=CH, O2CCH2,
CH2–CH=, O2CCH2CH2, other CH2, and CH3 in the ratios
1.00:1.00:1.50:1.00:7.00:0.75. The actual observed integra-
tion ratios were 1.00:1.01:1.52:0.98:7.03:0.69, respectively.
Thus, the experimental values were consistent with the theo-
retical ones.
The Z/E-configuration of these components can be deter-
mined by 13C NMR spectral analysis. The olefinic carbons of
the starting dioleate 5 (Z,Z-configuration) appear at 129.7 and
130.0 ppm. Component B shows two olefinic carbon peaks at
130.1 and 130.5 ppm. Considering that trans olefinic carbons
generally have higher chemical shifts than the corresponding
cis olefinic carbons, we have tentatively assigned compound
B the E,E-configuration. It has also been observed that the
13C NMR spectra of all of the oligomers formed in the reac-
tion show both cis (δ < 130.0 ppm) and trans (δ > 130.0 ppm)
4
SCHEME 4
tial work focused on an evaluation of the most widely used
catalysts today, Mo-based catalyst 1 and Ru-based catalyst 2.
Two representative dienes, 1,9-decadiene (3) and 5-hexenyl
4-pentenoate (4), were chosen as model monomers for
ADMET polymerization (Schemes 3 and 4). When diene 3
was allowed to react with catalyst 1, rapid evolution of ethyl-
ene (violent bubbling) was observed. However, the polymeri-
zation did not proceed to completion even after 4 d, and a liq-
1
uid was obtained. The H NMR spectrum of the product
showed two clearly separated peaks for the terminal olefinic
protons (δ = 4.97) and the newly formed internal olefinic pro-
tons (δ = 5.40). End-group analysis (16) based on the integra-
tion of these two types of olefinic protons indicated that the
number of the repeating units (Xn) in the polymer was only
three and the average M.W. (Mn) was 358. On the other hand,
when catalyst 2 was employed, the reaction started relatively
slowly. Although only a few bubbles were observed when cat-
alyst 2 was mixed with 3, the reaction proceeded more com-
pletely and solidified after 4 d. End-group analysis indicated
Xn = 142 and Mn = 15,620.
Similar results were observed for the reactions of diene 4. In
the presence of catalyst 1, a sticky oil was obtained. According
to end-group analysis, Xn was 20 and Mn was 2,840. When cat-
alyst 2 was used, the reaction afforded a very sticky oil, which
had substantially higher Xn (100) and Mn (14,200) values.
These results indicate that catalyst 1 is more reactive than
catalyst 2, but catalyst 2 affords polymers of higher M.W.
This can be explained by the high reactivity of catalyst 1 to-
ward moisture and oxygen and the fact that trace impurities
in the monomer will destroy the catalyst (16). On the other
hand, catalyst 2 is only mildly sensitive to oxygen and very
stable to water, alcohols, and most organic solvents (25).
Thus, catalyst 2 survives longer and affords polymers with
higher Xn and Mn. Catalyst 2 appears to be the better catalyst
for our ADMET polymerizations and was thus employed in
all subsequent work.
Polymerization of ethylene glycol dioleate. After the suc-
cessful ADMET polymerization of terminal dienes 3 and 4,
our attention turned toward the ADMET polymerization of
the internal diene ethylene glycol dioleate (5), easily prepared
in 68% yield from oleic acid and ethylene glycol. The
ADMET polymerization of dioleate 5 was conducted using
0.1 mol% catalyst 2 at 55°C under a vacuum for 24 h (Eq. 3).
JAOCS, Vol. 79, no. 5 (2002)

THE ADMET POLYMERIZATION OF SOYBEAN OIL
485
TABLE 3
TABLE 4
Results from MS Atmospheric Pressure Chemical Ionization (APCI)
Optimization of the Acyclic Diene Metathesis (ADMET)
Polymerization of Dioleate 5
and Gas Permeation Chromatography (GPC)a
Oligomer
FW
MS (APCI)
GPC (Mn)
Mn/FW
Catalyst
(mol%)
Time
(h)
Light-brown White, waxy
Entry
Vac/Ara
oil(w/w%)
solid (w/w%)
Dioleate
Dimer
Trimer
Tetramer
Pentamer
Hexamer
Heptamer
590
929
591 (MH)+
930 (MH)+
1268 (MH)+
1605 (MH)+
1946 (MH)+
2282 (M)+
1289
1796
2385
2996
3542
4443
6551
2.18
1.93
1.88
1.86
1.82
1.95
2.50
1
2
3
4
5
6
7
0.1
0.1
0.1
0.05
0.2
0.5
1.0
Vac
Vac
Ar
Vac
Vac
Vac
Vac
24
48
24
28
24
24
43
48
47
64
59
62
63
57
50
49
37
40
38
43
44
1267
1604
1945
2282
2621
2622 (MH)+
aMn, number average M.W.; FW, formula weight.
aFor abbreviation see Table 1.
olefinic carbon peaks, where the latter peaks are much more
intense than the former ones. This indicates that the carbon- fraction was further partitioned by flash chromatography, and
carbon double bonds in these oligomers have both Z- and E- eight components were isolated (Table 2).
configurations, with the E-configuration predominating.
The structural assignments for these components (Table 2,
1
1
As shown in Table 3, the results from MS (APCI) match entries 1–8) are based on H NMR spectral data. The H
the corresponding formula weights (FW). However, there are NMR spectra of these components are similar to the starting
differences between the GPC results and the FW. Presumably, trioleate, differing only in the peak integrations. Analysis of
this is due to the difference in the hydrodynamic volume of the peak integration data provides information about the
the polyester samples and the polystyrene standard utilized structure of these components. Our attention was focused on
1
(16). Similar discrepancies were observed in all of the GPC the ratio of the H NMR spectral peaks corresponding to
experiments, including the oligomers prepared from glyceryl OCH2CHCH2O to the terminal CH3 protons. This ratio is de-
trioleate and soybean oil to be discussed later. As shown in fined as r. If the trioleates link only with each other in a
Table 3, entry 1, the Mn/FW ratio for the dioleate is 2.18, and straight chain with no intramolecular cyclization, the value of
the other ratios are similar to this value. This suggests that the x in Equation 4 should be equal to (n − 1). Then the ratio r
Mn is approximately linear with FW and Mn can be corrected can be described as 4n/[9n − 6(n − 1)], i.e., 4n/(3n + 6). Ac-
by dividing by a factor of approximately 2.
cording to this equation, the theoretical values of r for the
The polymerization reaction has also been run under a vari- acyclic oligomers, from monomer to octamer, are recorded in
ety of reaction conditions in an effort to improve the yield of column 4 of Table 2 (entries 1–8, n = 1–8). A comparison of
solid product and minimize the amount of catalyst 2 used the experimental values of r (Table 2, column 3) and the the-
(Table 4). It has been shown that the presence of a vacuum is oretical ones (Table 2, column 4) suggests that components 1
necessary for the reaction to produce good yields of solid prod- to 5 can be tentatively assigned as isomerized trioleate, dimer,
uct (compare entries 1 and 3). The smallest amount of catalyst trimer, tetramer, and pentamer, respectively. The 13C NMR
used in the reaction was 0.05 mol% (entry 4), and the reaction spectrum of component 1 shows new peaks for the olefinic
still afforded a fair yield of solid product (40%). However, carbons (δ = 130.2, 130.5 ppm), which presumably corre-
when more catalyst was employed, the yields of solid product spond to the E-isomer, suggesting that component 1 is simply
did not increase accordingly (entries 5–7). The best yield was the isomerized E,E,E-trioleate. The olefinic carbons in com-
obtained when 0.1 mol% of catalyst 2 was used (entry 1).
pounds 2–8 show intense peaks at δ = 130.2–130.6 ppm. This
Polymerization of glyceryl trioleate and glyceryl tri- indicates that the carbon-carbon double bonds in these com-
linoleate. Glyceryl trioleate was readily prepared in 51% pounds are predominantly of the E-configuration. This phe-
yield from oleic acid and glycerol. Polymerization of the tri- nomenon is consistent with observations made during our
oleate was examined by employing the best conditions devel- study of the oligomers formed in the ADMET polymerization
oped for the polymerization of the dioleate 5 (Eq. 4). The re- of the ethylene glycol dioleate 5.
sulting crude product (a very sticky oil) was dissolved in
Note that there is a significant discrepancy between the ex-
CH2Cl2 and then poured into an excess of MeOH (workup perimental and theoretical values of r for components 6–8
procedure A). Two fractions, a MeOH/CH2Cl2-soluble frac- (Table 2, compare columns 3 and 4 of entries 6–8). This sug-
tion (32%, w/w) and a MeOH/CH2Cl2-insoluble fraction gests the possibility of intramolecular cyclization. If only one
(67%, w/w), were isolated. The MeOH/CH2Cl2-insoluble intramolecular cyclic ring is formed in the oligomer, the value
0.1 mol% of catalyst 2
CH3(CH2)7CH=CH(CH2)7CO2
55°C, vacuum
n
—O2C(CH2)7CH=CH(CH2)7CH3
24 h
CH3(CH2)7CH=CH(CH2)7CO2
[4]
CH3(CH2)3CH =CH(CH2)7CO2
CH3(CH2)3CH =CH(CH2)7CO2
—O2C(CH2)7CH= CH(CH2)3CH3 + xCH3(CH2)7CH=CH(CH2)7CH3
n
JAOCS, Vol. 79, no. 5 (2002)

486
Q. TIAN AND R.C. LAROCK
TABLE 5
for 24 h. A brown rubber was obtained in 87% yield. This
rubbery material did not dissolve in organic solvents, such as
CH2Cl2, chloroform, toluene, or benzene. Under the same
conditions, the polymerization of glyceryl trilinoleate also
produced a brown rubber, which was insoluble in organic sol-
vents. We suspect that these rubbery materials are highly
cross-linked polymers.
In summary, our model system studies suggest that poly-
merization of the dioleate and trioleate follow the expected
ADMET mechanism. Since dioleate 5 or its oligomer is a lin-
ear molecule with a long chain, the intermolecular ADMET
GPC Results for the Monocyclic Oligomers and the Approximate
Number of Repeating Unitsa
Entry
Component
Mn
Mw
PDI
n
1
2
3
4
Trioleate
1172
5375
8182
1518
7448
11478
27963
1.29
1.38
1.40
1.60
1
6
7
8
6
10
21
17459
aThe value of n is calculated as n = Mn/(1.32 × 632). Mw , weight average
M.W.; PDI, polydispersity. For other abbrevation see Table 3.
of x in Equation 4 is expected to be n; thus, r can be calcu- reaction is more favorable than the intramolecular cyclization.
lated as 4n/(9n − 6n), i.e., 1.33. This indicates that all mono- Thus, the polymerization of dioleate 5 affords a series of
cyclic oligomers have the same value of r, regardless of the oligomers, from dimer to heptamer, with the dimer and trimer
number of repeating units (n) in the polymer or the size of the predominating. On the other hand, the trioleate and its
cyclic ring. Since the experimental values of r (Table 2, en- oligomer possess structures like a dendrimer. This allows an
tries 6–8) are close to 1.33, we have tentatively assigned com- olefinic group in an appropriate position of the same molecule
ponents 6–8 as monocyclic oligomers. For component 8, the to cyclize through the ADMET process. The polymerization
experimental value of r is 1.43. This is even higher than 1.33 of the trioleate produced monocyclic oligomers with much
and indicates the possibility of bicyclic oligomers in compo- higher M.W. owing to possible cross-linking polymerization.
nent 8, since the r value of the bicyclic oligomers can be cal-
culated as 4n/(3n − 6), which is larger than 1.33.
Polymerization of soybean oil. Encouraged by the success of
our model system studies, the ADMET polymerization of soy-
The GPC results from analysis of the starting trioleate and bean oil was examined. New Horizons soybean oil produced by
the presumed monocyclic components 6–8 are summarized in Pioneer Hi-Bred, Inc. (Des Moines, IA) was employed in the
Table 5. A discrepancy between the FW of the trioleate (884) following reactions. The results are summarized in Table 1. De-
and its Mn (1172) is observed (entry 1) and the ratio of Mn over pending on the reaction conditions, two types of products, a
the calculated FW is 1.32. In analogy to our previous results sticky oil and a rubbery material, have been observed. Accord-
from the polymerization of the dioleate 5, which suggested ingly, two efficient workup procedures, procedures A and B,
that Mn is approximately linear with FW, we have assumed have been developed for these two types of materials. In proce-
that this linear relationship also exists for the trioleate system. dure A, which is applicable to the sticky oil, the crude product
Thus, Mn in Table 5 is corrected by a factor of 1.32. The cor- is dissolved in CH2Cl2 and then poured into an excess of
responding numbers of repeating units can be approximately MeOH. The MeOH solution and the residual material are col-
calculated from these corrected Mn. According to Equation 4, lected as fraction A and fraction B, respectively. If the crude
the M.W. of the monocyclic oligomers should be the mass of product is a rubbery material, procedure B is utilized for the
the reactant (884 × n) minus the mass of 9-octadecene (252 × workup. In procedure B, the crude product is first worked up
n); so the number average M.W. is 632 × n. Then n = Mn/(1.32 utilizing procedure A. The MeOH/CH2Cl2-soluble fraction is
× 632). The results are summarized in Table 5. The average recorded as fraction A and the residual rubbery material is fur-
number of repeating units of component 6 is thus calculated ther partitioned by Soxhlet extraction. The CH2Cl2 solution is
as 6. This appears to be reasonable, since the previous compo- collected as fraction B, and the remaining rubbery material is
nent (component 5) was assigned as a pentamer (n = 5). As ex- collected as fraction C. A “blank” workup experiment in which
pected, the average number of repeating units of component 7 soybean oil was processed using procedure A has shown that
is even higher than that of component 6 (n ≅ 10, Table 5, entry soybean oil is recovered quantitatively in fraction A.
1
3). It is very interesting that component 8, which is the major
As proven in our model system studies, H NMR spec-
oligomer formed in the reaction, has a high average repeating troscopy can provide valuable information about the struc-
unit number. The significantly higher degree of polymeriza- tural makeup of the oligomers. Again, r, the ratio of the
tion of the trioleate (n ≅ 6, 10, 21) than that of the dioleate 5 OCH2CHCH2O protons over the terminal CH3 protons, has
(n = 2–4) suggests that there has been additional polymeriza- served as an important parameter in our understanding of the
tion by cross-linking. This is to be expected, since the trioleate, structure of the oligomers. The 1H NMR spectra of soybean oil
which has three double bonds, may behave as a trifunctional and the fractions A and B obtained from the polymerization of
monomer to produce dendrimers. In the ideal dendrimer case, soybean oil all have the same peaks, but the integrations differ.
after m generations, n can be calculated as n = 1 + 3m. Thus, a For soybean oil itself, r is equal to 0.56. This is measured from
second-generation dendrimer affords n = 10, and the third- the 1H NMR spectrum of the soybean oil. The value of r for
generation dendrimer will yield n = 28.
fraction A ranged from 0.21 to 0.47 depending on the reaction
Other experiments also confirm the existence of cross- conditions. This is lower than that of soybean oil and consis-
linking polymerization. The polymerization of the trioleate tent with our expectation that fraction A contains unreacted or
has been examined in the presence of 1.6 mol% of catalyst 2 isomerized soybean oil and long-chain alkene by-product.
JAOCS, Vol. 79, no. 5 (2002)

THE ADMET POLYMERIZATION OF SOYBEAN OIL
487
O
C
O
O
O
C
catalyst 2
n
6
[5]
O
C
O
C
O
O
+ (n − 1) H2C=CH2
CH2
H2C
n
In fact, the GC–MS results obtained from the analysis of yield of fraction B was not as good as that obtained when the
fraction A in Table 1, entry 4, showed peaks with masses of reaction was run under vacuum (entry 12).
102, 168, 208, 220, 250, 290, and 330. All of these masses
The reaction with 0.01 mol% of catalyst 2 failed (entry 14).
can be assigned to the alkenes expected from the ADMET In entry 15, without any catalyst, soybean oil was quantitatively
reaction of the TG in soybean oil. For example, mass 208 recovered in fraction A after the reaction was run under vac-
presumably comes from the alkene CH3(CH2)7CH=CH– uum at 55°C for 212 h. This indicates that no thermal polymeri-
CH2CH=CHCH2CH3. The mass of CH3CH2CH=CH– zation occurred at 55°C and no oil was lost due to the vacuum.
CH2CH=CHCH2CH=CHCH2CH=CHOCH2CH3 is 220. This
In summary, as observed in the ADMET polymerization
product is expected to be formed by metathesis of the soy- of the dioleate 5, 0.1 mol% of catalyst 2 appears to give the
bean oil and the ethyl vinyl ether used to quench the reaction. best results. Under these conditions, the reaction affords ei-
The mass 250 can be assigned as CH3(CH2)7CH=CH– ther a rubbery material at long reaction times (entry 11) or a
CH2CH=CH(CH2)4CH3.
sticky oil at short reaction times (entry 12).
GC–MS analysis of fraction A from Table 1, entry 11,
Copolymerization of soybean oil. Di-(4-pentenyl)1,4-ben-
showed only three peaks with masses of 220, 250, and 290. zenedicarboxylate (6) was prepared according to a literature
The other low M.W. alkenes expected to be formed may have procedure (16). The ADMET polymerization of 6 was
been removed by vacuum. The value of r for fraction B was achieved using 0.5 mol% of catalyst 2. End-group analysis
found to be higher than that of soybean oil. This indicates the indicated that Mn = 3718 and Xn = 13 (Eq. 5).
formation of oligomers. The rubbery materials of fraction C
are presumably highly cross-linked polymers.
The copolymerization of soybean oil and diester 6 was in-
vestigated. In the presence of 0.5 mol% of catalyst 2, a 50:50
In the early stages of our investigation, a relatively high (w/w) mixture of soybean oil and diester 6 produced a
amount of catalyst 2 (1.5%) was employed and the reaction copolymer with 41% weight recovery. 1H NMR spectral data
afforded a brown rubber, which did not dissolve in any com- indicated that the molar ratio of diester 6 to soybean oil in the
mon organic solvent (Table 1, entry 1). Soybean oil purified copolymer was 4.37, which was slightly higher than that in
by pretreatment with CaH2, MgSO4, or Ac2O gave the same the reactant mixture (3.09). This difference is presumably due
result as the original soybean oil. Thus, soybean oil taken di- to a difference in the relative reactivities of diester 6 and soy-
rectly from the bottle was used in the following reactions bean oil. As a terminal diene, diester 6 should be more reac-
without any further purification.
tive in ADMET polymerization than soybean oil, which con-
The ADMET polymerization reactions proceeded very tains only internal carbon-carbon double bonds. Similar re-
slowly at room temperature, and long reaction times (5–10 d) sults have also been observed in the copolymerization of
were required (Table 1, entries 1–4). The reaction time can be soybean oil and norbornene. This latter reaction is currently
reduced substantially by increasing the temperature to 55°C being studied further since it produces interesting coatings.
(entry 5). The effect of vacuum was also examined. Under vac-
uum, the reaction afforded an insoluble rubber (entries 1–3 and of 1,9-decadiene and 5-hexenyl 4-pentenoate have suggested
5), but no rubbery material was obtained under Ar (entry 4). that Grubbs’ ruthenium catalyst 2 is more efficient than
In conclusion, our studies of the ADMET polymerization
Less catalyst also was employed. When 0.2 mol% of cata- Schrock’s molybdenum catalyst 1. In the presence of 0.1
lyst 2 was employed, the reaction at room temperature af- mol% of catalyst 2, the ADMET polymerization of ethylene
forded a brown oil after 240 h either under vacuum or Ar glycol dioleate afforded isomerized E,E-dioleate and a series
(entries 7 and 8). When the reaction was run with 0.2 mol% of oligomers ranging from dimer to heptamer. To our knowl-
catalyst 2 at 55°C, a rubbery material was obtained under vac- edge this is the first example of an internal diene participating
uum (entry 9), but a brown oil was obtained under Ar (entry efficiently in ADMET polymerization. Under the same con-
10). A comparison of entries 3 and 7 indicates that higher ditions, glyceryl trioleate underwent ADMET polymerization
amounts of catalyst favor the formation of rubbery materials.
to produce dimer, trimer, tetramer, pentamer, and monocyclic
In the presence of 0.1 mol% of catalyst 2 under vacuum, a oligomers, with high-M.W. monocyclic oligomers predomi-
rubbery material was obtained after 192 h (entry 11). How- nating. We also succeeded in the ADMET polymerization of
ever, if the reaction was stopped earlier (24 h), a brown oil soybean oil. A variety of materials, from sticky oils to rub-
was obtained in 90% yield (entry 12). The analogous reaction bers, were prepared from soybean oil. These materials hold
under Ar afforded a similar oily product (entry 13), but the industrial promise and may well be biodegradable.
JAOCS, Vol. 79, no. 5 (2002)

488
Q. TIAN AND R.C. LAROCK
13. Wagener, K.B., J.G. Nel, J. Konzelman, and J.M. Boncella,
ACKNOWLEDGMENTS
Acyclic Diene Methathesis Copolymerization of 1,5-Hexadiene
and 1,9-Decadiene, Macromolecules 23:5155–5157 (1990).
14. Wagener, K.B., J.M. Boncella, and J.G. Nel, Acyclic Diene
Methathesis (ADMET) Polymerization, Macromolecules 24:
2649–2657 (1991).
15. Bauch, C.G., K.B. Wagener, and J.M. Boncella, Acyclic Diene
Metathesis (ADMET) Polymerization. Synthesis of an Unsatu-
rated Polyester, Makromol. Chem., Rapid Commun. 12:
413–417 (1991).
16. Paton, J.T., J.M. Boncella, and K.B. Wagener, Acyclic Diene
Methathesis (ADMET) Polymerization. The Synthesis of Un-
saturated Polymers, Macromolecules 25:3862–3867 (1992).
17. Davidson, T.A., K.B. Wagener, and D.B. Priddy, Polymeriza-
tion of Dicyclopentadiene: A Tale of Two Mechanisms, Macro-
molecules 29:786–788 (1996).
18. Wolfe, P.S., and K.B. Wagener, Acyclic Diene Metathesis
(ADMET) Polymerization: The Investigation of Dienes Con-
taining the Boronate Functionality, Polym. Prepr. 37:439–440
(1996).
19. Walba, D.M., P. Keller, R. Shao, N.A. Clark, M. Hillmyer, and
R.H. Grubbs, Main-Chain Ferroelectric Liquid Crystal
Oligomers by Acyclic Diene Metathesis Polymerization, J. Am.
Chem. Soc. 118:2740–2741 (1996).
20. Schwab, P.R., H. Grubbs, and J.W. Ziller, Synthesis and Appli-
cations of RuCl2(=CHR′)(PR3)2: The Influence of the Alkyl-
idene Moiety on Metathesis Activity, J. Am. Chem. Soc. 118:
100–110 (1996).
21. Wolfe, P.S., F.J. Gómez, and K.B. Wagener, Metal-Containing
Polymers Synthesized via Acyclic Diene Metathesis: Polycar-
bostannanes, Macromolecules 30:714–717 (1997).
22. Erhan, S.Z., and M.O. Bagby, Polymerization of Vegetable Oils
and Their Uses in Printing Inks, J. Am. Oil Chem. Soc. 71:
1223–1226 (1994).
23. Johnson, L.A., and D.T. Myers, Industrial Uses for Soybeans,
Practical Handbook of Soybean Processing and Utilization,
edited by D.R. Erickson, AOCS Press, Champaign, 1995, p.
380.
We thank the Iowa Soybean Promotion Board for support of this re-
search. Thanks also go to Professor Kenneth B. Wagener (Depart-
ment of Chemistry and Center for Macromolecular Science and En-
gineering, University of Florida) for providing valuable suggestions
and donating catalysts 1 and 2. We also thank Professor Valerie
Sheares at Iowa State University for assistance with the GPC exper-
iments. The soybean oil used in this research was provided by Pro-
fessor Earl Hammond (Department of Food Science and Human Nu-
trition, Iowa State University).
REFERENCES
1. Ivin, K.J., Olefin Metathesis, Academic Press, London, 1983.
2. Amass, A.J., Metathesis Polymerization: Chemistry, Compr.
Polym. Sci. 4:109–134 (1989).
3. Van Dam, P.B., M.C. Mittelmeijer, and C. Boelhouwer,
Metathesis of Unsaturated Fatty Acid Esters by a Homogeneous
Tungsten Hexachloride–Tetramethyltin Catalyst, J. Chem. Soc.,
Chem. Commun.:1221–1222 (1972).
4. Van Dam, P.B., M.C. Mittelmeijer, and C. Boelhouwer, Homo-
geneous Catalytic Metathesis of Unsaturated Fatty Esters: New
Synthetic Method for Preparation of Unsaturated Mono- and Di-
carboxylic Acids, J. Am. Oil Chem. Soc. 51:389–392 (1974).
5. Boelhouwer, C., and J.C. Mol, Metathesis Reactions of Fatty
Acid Esters, Prog. Lipid Res. 24:243–267 (1985).
6. Verkujilen, E., and C. Boelhouwer, Formation of Cyclohexa-
1,4-diene by Metathesis of Linoleic and Linolenic Esters, J.
Chem. Soc., Chem. Commun.:793–794 (1974).
7. Kohashi, H., and T.A. Foglia, Metathesis of Methyl Oleate with
a Homogeneous and a Heterogeneous Catalyst, J. Am. Oil
Chem. Soc. 62:549–554 (1985).
8. Mandelli, D., M.J.D.M. Jannini, R. Buffon, and U. Schuchardt,
Ethenolysis of Esters of Vegetable Oils: Effect of B2O3 Addi-
tion to Re2O7/SiO2·Al2O3-SnBu4 and CH3ReO3/SiO2·Al2O3
Metathesis Catalysts, J. Am. Oil Chem. Soc. 73:229–232 (1996).
9. Erhan, S.Z., M.O. Bagby, and T.C. Nelsen, Drying Properties of
Metathesized Soybean Oil, J. Am. Oil Chem. Soc. 74:703–706
(1997).
10. Refvik, M.D., R.C. Larock, and Q. Tian, Ruthenium-Catalyzed
Metathesis of Vegetable Oils, J. Am. Oil Chem. Soc. 76:93–98
(1999).
11. Finkel’shtein, E.S., V.I. Bykov, and E.B. Portnykh, The Olefin
Metathesis Reaction—A Versatile Tool for Fine Organic Syn-
thesis, J. Mol. Catal. 76:33–52 (1992).
24. Duralski, A.A., P.R. Spooner, and A. Watts, Synthesis of Opti-
cally Active Polyunsaturated Diacylglycerols, Tetrahedron Lett.
30:3585–3588 (1989).
25. Nguyen, S.T., and R.H. Grubbs, Syntheses and Activities of
New Single-Component Ruthenium-Based Olefin Metathesis
Catalysts, J. Am. Chem. Soc. 115:9858–9859 (1993).
12. Nel, J.G., K.B. Wagener, J.M. Boncella, and R.P. Duttweiler,
Acyclic Metathesis Polymerization. The Synthesis of Highly
trans Polyoctenamer, Polym. Prepr. 30:283–284 (1989).
[Received September 7, 2000; accepted February 7, 2002]
JAOCS, Vol. 79, no. 5 (2002)