Lactic acid and ricinoleic acid based copolyesters

Article abstract of DOI:10.1021/ma0503918

Copolyesters based on purified ricinoleic (RA) and lactic (LA) acids with different RA:LA ratios were synthesized by thermal polycondensation and by transesterification of high molecular weight poly(lactic acid) (PLA) with ricinoleic acid and repolyesterification. Thermal polycondensation resulted in random P(LA-RA) copolyesters of molecular weights between 2000 and 8000 with the polymers containing 20% or more RA were liquid at room temperature. Transesterification of high molecular weight PLA with pure ricinoleic acid and repolymerization of those oligomers by condensation resulted in multiblock P(PLA-RA) copolyesters of molecular weights between 6000 and 14000. Polymers containing 50% RA were liq uid at room temperature. ¹H NMR spectroscopy analysis coupled with information from DSC allowed de ;ermination of the polymer structure. Polymers prepared by thermal polycondensation are random cope lymers (h > 1), while the copolymers prepared by transesterification have a multiblock character (h < 1). The LA number-average sequence length (L_LA) decreased from 12 to 4 for the LA-RA 9:1 and 5:5 copolymers prepared by thermal polycondensation and from 50 to 17 for the corresponding LA-RA copolymers prepared by transesterification. Thermal analysis by DSC revealed crystalline structure for polyester synthesized by transesterification. For polyesters synthesized by random condensation on y P(LA-RA) 90:10 w/w contained crystalline domains.

Full text of DOI:10.1021/ma0503918

Macromolecules 2005, 38, 5545-5553
5545
Lactic Acid and Ricinoleic Acid Based Copolyesters
Raia Slivniak and Abraham J. Domb*
Department of Medicinal Chemistry and Natural Products, School of Pharmacy, Faculty of Medicine,
The Hebrew University of Jerusalem, 91120 Jerusalem, Israel
Received February 23, 2005
ABSTRACT: Copolyesters based on purified ricinoleic (RA) and lactic (LA) acids with different RA:LA
ratios were synthesized by thermal polycondensation and by transesterification of high molecular weight
poly(lactic acid) (PLA) with ricinoleic acid and repolyesterification. Thermal polycondensation resulted
in random P(LA-RA) copolyesters of molecular weights between 2000 and 8000 with the polymers
containing 20% or more RA were liquid at room temperature. Transesterification of high molecular weight
PLA with pure ricinoleic acid and repolymerization of those oligomers by condensation resulted in
multiblock P(PLA-RA) copolyesters of molecular weights between 6000 and 14000. Polymers containing
50% RA were liquid at room temperature. 1H NMR spectroscopy analysis coupled with information from
DSC allowed determination of the polymer structure. Polymers prepared by thermal polycondensation
are random copolymers (h > 1), while the copolymers prepared by transesterification have a multiblock
character (h < 1). The LA number-average sequence length (LLA) decreased from 12 to 4 for the LA-RA
9:1 and 5:5 copolymers prepared by thermal polycondensation and from 50 to 17 for the corresponding
LA-RA copolymers prepared by transesterification. Thermal analysis by DSC revealed crystalline
structure for polyester synthesized by transesterification. For polyesters synthesized by random
condensation only P(LA-RA) 90:10 w/w contained crystalline domains.
1. Introduction
with other comonomers, leading to the incorporation of
units disturbing the crystallization ability of the PLA
segments. For example, glycolide, ꢀ-caprolactone, δ-val-
erolactone, 1,5-dioxepan-2-one (DXO) and trimethylene
carbonate (TMC) are frequently used as comonomers in
order to change thermal properties of resulting PLA
copolymers. The Tg of PLA copolymer consisting of 28%
DXO was reduced by 35 °C compared to a pure PLLA,
and the crystallinity was also significantly lower. A
linear decrease in Tg was also found for copolymers with
more then 10% w/w TMC. A 50/50 copolymer of L-lactide
and ꢀ-caprolactone has been shown to have Tg ) -15
°C indicating a continuous amorphous phase.13-16
Fatty acids are suitable candidates for the prepara-
tion of biodegradable polymers,2,3,17,18 as they are natu-
ral body components and they are hydrophobic, and thus
they may retain an encapsulated drug for longer time
periods when used as drug carriers. However, most fatty
acids are monofunctional and cannot serve as monomers
for polymerization. Ricinoleic acid is a common C18 fatty
acid with a cis-configured double bond in the ninth
position and a hydroxyl group in the 12th position (cis-
12-hydroxyoctadeca-9-enoic acid). It is produced from
the hydrolysis of castor oil.19
The objective of this study is to incorporate ricinoleic
acid in lactide based polymers for the purpose of altering
its physical properties. The bifunctionality of ricinoleic
acid allows its polymerization without any modifica-
tions. Copolymerization of PLA with ricinoleic acid may
result in polymers with desired properties such as
pliability, hydrophobicity and softness. Ricinoleic acid
hydrocarbon chain, cis configuration, and side chain
may result in steric hindrance of the polymer to yield
soft or even liquid polymers. In addition, the presence
of a double bond in ricinoleic acid provides additional
synthetic opportunities including cross-linking, hy-
droxylation, bromination, and other common modifica-
tions of double bonds.
Biodegradable polyesters are useful materials for
controlled drug delivery.1,2 They hydrolyze to hydroxy
acid monomers when placed in aqueous medium.3-5 The
biodegradable polymers that have been most intensively
investigated are aliphatic polyesters of both natural and
synthetic origins. Polyesters can be synthesized by
polycondensation of hydroxy acids or by ring-opening
polymerization of cyclic esters (lactones). A wide range
of monomers have been used to produce biodegradable
polyesters. Their polymerizations can be carried out
either in the bulk or in solution. The most useful
monomers used for polycondensation are lactic, glycolic,
hydroxybutyric, and hydroxycaproic acids. Polyesters of
glycolic and lactic acids are the main group of interest
due to their long history of safety.5-8
Lactic acid based polymers have been prepared by
direct polycondensation, ring-opening polymerization
(ROP), grafting, chain extension, or transesterification.9
Lactic acid can be condensed with other hydroxy acids
such as 6-hydroxycaproic acid, glycolic acid, and hy-
droxybutyric acid or in the presence of diols, diacids,
and diamine. Direct condensation usually resulted in
low molecular weight copolymers that can then be
further linked to yield high molecular weight polymers.
In the second step, linking molecules such as diisocy-
anates, bis(amino-ethers), phosgene, phosphate, and
anhydrides takes place.10-12
Enantiomerically pure PLA is a semicrystalline poly-
mer with Tg of about 55 °C and Tm of about 180 °C. The
degree of crystallinity and melting temperature of PLA
polymers can be reduced by random copolymerization
* Corresponding author. Telephone: 972-2-6757573. Fax: 972-
2-6757629. E-mail: adomb@md.huji.ac.il. Affiliated with the David
R. Bloom Center for Pharmacy and the Alex Grass Center for
Synthesis and Drug Design at The Hebrew University of Jerusa-
lem.
10.1021/ma0503918 CCC: $30.25 © 2005 American Chemical Society
Published on Web 06/02/2005

5546 Slivniak and Domb
Macromolecules, Vol. 38, No. 13, 2005
conducted in 0.1 M phosphate buffer (pH 7.4) at 37 °C with a
Previous studies in our laboratory focused on the
synthesis of ricinoleic acid based polyanhydrides.18,20
Polyanhydrides synthesized from ricinoleic acid maleate
or succinate and sebacic acid possessed desired physi-
cochemical properties such as low melting temperature,
hydrophobicity, and pliability, in addition to biocom-
patibility and biodegradability. The polymers were
synthesized by melt condensation to yield film-forming
polymers with molecular weights exceeding 100 000. In
another study, fatty acid esters of ricinoleic acid were
used as chain terminators of polyanhydrides based on
sebacic acid.3
In our previous article we reported on the synthesis
of ricinoleic acid lactones and their homopolymerization
and copolymerization with lactide by ring-opening po-
lymerization. A systematic study on the synthesis and
characterization of ricinoleic acid lactones and ricinoleic
acid-co-lactic acid polyesters was reported.21
The present study is a continuation of our efforts to
synthesize new biodegradable polymers based on rici-
noleic acid. The purpose of this study has been to
synthesize ricinoleic acid based copolyesters with lactic
acid using different polymerization techniques and
determine the copolymer chain structure as a function
of ricinoleic acid content and method of synthesis.
Various techniques have previously been used to
study and characterize polyesters and polyanhydrides.22
Yet, quantitative correlation between physical proper-
ties and the 1H NMR spectroscopy was reported mainly
for polyanhydrides.23 A systematic study on the syn-
thesis and characterization of ricinoleic acid-co-lactic
acid polyesters is reported.
1
constant shaking of 100 rpm. H NMR and 13C NMR spectra
(in CDCl3) were recorded on a Varian 300 and 500 MHz
spectrometers using TMS as internal standard (Varian Inc.,
Palo Alto, CA). Optical rotations of polymers were determined
by Optical Activity LTD polarimeter (Cambridgeshire, Eng-
land) in 10 mg/mL polymer in CHCl3 solution. Viscometry of
the polymers in dichloromethane was measured in Cannon-
Ubbelohde 75 µm dilution viscometer. Afflux times were
measured at four concentrations at 25 °C; the data were
analyzed by standard methods.
2.3. Ricinoleic Acid Purification. Crude ricinoleic acid
(100 g) was converted to ricinoleic acid methyl ester by
bubbling gaseous HCl, throw methanol solution of ricinoleic
acid at 0 °C. Ricinoleic acid methyl ester was extracted with
hexanes, washed with DDW and with 10% (w/v) NaHCO3,
treated with anhydrous MgSO4, filtered and evaporated to
dryness. Yield ) 102 g of ricinoleic acid methyl ester (∼97%).
TLC (10% acetone in hexanes) showed two major spots: 1. Rf
) 0.75; methyl esters of non-hydroxy fatty acids (i.e., oleic,
linoleic, stearic, etc.), and 2. Rf ) 0.375; methyl ester of
ricinoleic acid. The crude ricinoleic acid methyl ester was
diluted with hexanes and purified by silica gel chromatogra-
phy. Ricinoleic acid methyl ester was eluted with solution of
10% acetone in hexanes. Fractions containing different prod-
ucts were separately combined and evaporated to dryness
under reduced pressure. A 55 g sample of pure ricinoleic acid
methyl ester [R]20 ) +3.4° was collected.
D
1H NMR (CDCl3, #16-#25): 5.59-5.37 (2H, m, C9-10,
-CHdCH-), 3.62 (3H, s, -CO2CH3), 3.599 (1H, m, C12 HC-
O-), 2.322 (2H, t, C2 -CH2), 2.202 (2H, t, C11 -CH2), 2.053
(2H, m, C8 -CH2), 1.59 (2H, m, C3 -CH2), 1.438 (2H, m, C13
-CH2), 1.296 (16H, m, C4-7 and C14-17), and 0.854 (3H, t,
C18 -CH3) ppm.
To obtain pure ricinoleic acid, ricinoleic acid methyl ester
was hydrolyzed in alkaline water: methanol 1:1 solution. Then
the mixture was diluted with DDW and acidified with con-
centrated HCl solution to a final pH of 2. Pure ricinoleic acid
was extracted with ethyl acetate, treated with anhydrous
2. Experimental Section
2.1. Materials. Crude ricinoleic acid was purchased from
Acros (85% pure) (Geel, Belgium), L-Lactic acid (L-LA) and DL-
lactic acid (DL-LA) were purchased from J. T. Baker (Deventer,
The Netherlands). D-Lactic acid was prepared from the hy-
drolysis of D-lactide in water; D-lactide was purchased from
Purac Biochem (Gorinchem, The Netherlands). CDCl3, for
NMR, was purchased from Sigma-Aldrich (Rehovot, Israel).
All solvents and salts were analytical grade from Aldrich or
Biolab (Jerusalem, Israel).
MgSO4, filtered and evaporated to dryness, [R]20 ) +3.8°.
D
1H NMR (CDCl3) showed the disappearance of the peak at
3.62 which confirmed the complete removal of the methyl ester
protecting group.
1H NMR (CDCl3, pure ricinoleic acid): 5.57-5.38 (2H, m,
C9-10, -CHdCH-), 3.609 (1H, m, C12 HC-O-), 2.309 (2H,
t, C2 -CH2), 2.194 (2H, t, C11 -CH2), 2.01 (2H, m, C8 -CH2),
1.603 (2H, m, C3 -CH2), 1.446 (2H, m, C13 -CH2), 1.291 (16H,
m, C4-7 and C14-17), and 0.862 (3H, t, C18 -CH3) ppm.
2.4. Copolymer Synthesis by Thermal Polycondensa-
tion. Low molecular weight polyesters, PRA, L- and D-PLA,
P(L-LA:RA), P(D-LA:RA) with different LA:RA (w/w) ratios
were prepared by two-step thermal polycondensation according
to the following procedure.
2.2. Instrumentation. IR spectra were performed on
monomer and polymer samples cast on NaCl plates from CH2-
Cl2 solutions on Bruker (Vector 22 System FT-IR). UV spectra
were taken on a Kontron Instruments Uvicon model 930
(Msscientific, Berlin, Germany). Thermal analysis was deter-
mined on a Mettler TA 4000-DSC differential scanning
calorimeter (Mettler-Toledo. Schwerzzenbach, Switzerland),
calibrated with Zn and In standards, at a heating rate of 10°C/
min under nitrogen atmosphere. Melting temperatures of the
co-polyesters was determined also by Fisher Scientific melting
point apparatus. Molecular weights of the co-polyesters were
estimated on a gel permeation chromatography (GPC) system
consisting of a Waters 1515 Isocratic HPLC Pump, with 2410
Refractive Index detector (RI) (Waters, MA), a Rheodyne
(Coatati, CA) injection valve with a 20 µL loop. Samples were
eluted with chloroform through a linear Styrogel column, 500
Å-pore size (Waters, MA) at a flow rate of 1 mL/min. The
molecular weights were determined relative to polystyrene
standards (Polyscience, Warrington, PA) with a molecular
weight range of 500 to 20 000 using BREEZE 3.20 version,
copyright 2000 Waters corporation computer program. The
lactic and ricinoleic acids release was determined by HPLC
using C18 reverse-phase column (LichroCart 250-4, Lichro-
spher 100, 5µm). Lactic acid was eluted with a solution of 0.1%
H3PO4 in DDW at a flow rate of 1 mL/min and UV detection
at 210 nm. Ricinoleic acid was eluted with a solution of
acetonitrile:0.1% H3PO4 in DDW 65:35 v/v, at flow rate of 1.4
mL/min and UV detection at 210 nm. The hydrolysis was
First, a 250 mL round-bottomed flask, equipped with a
Dean-Stark reflux condenser and CaCl2 drying tube, was
charged with pure ricinoleic acid and lyophilized lactic acid
in appropriate ratios (total amount of both acids was 20 g) and
150 mL of toluene.
The acid mixture was dried overnight with refluxing toluene
to remove water traces, then toluene was removed and the
temperature was raised gradually to 180 °C. The acids were
condensed for 3 h. In the second step, the temperature was
decreased back to 150 °C and the reaction flask was connected
to an oil pump where the condensation was continued under
a vacuum of 0.3 mmHg for additional 12 h. Each step was
followed by GPC analysis of samples to determine the molec-
ular weight of the forming polymers at each time period. All
polymers were characterized by GPC, 1H NMR, IR, DSC, mp,
a Cannon-Ubbelohde 75 dilution viscometer, and specific
optical rotation.
1H NMR (CDCl3, P(LA-RA) 60:40, δ): 5.45-5.30 (2H, m,
C9-10, -CHdCH-), 5.20-5.02 (1H, q, CH-CH3, LA), 4.94-
4.86 (1H, m, C12 HC-O-), 2.38-2.24 (2H, m, C2 -CH2, and
2H, m, C11 -CH2), 2.01 (2H, m, C8 -CH2), 1.68-1.50 (2H,

Macromolecules, Vol. 38, No. 13, 2005
Lactic Acid and Ricinoleic Acid Based Copolyesters 5547
Table 1. Physical Properties of RA-LA Copolyesters, Synthesized by Random Polycondensation of the Acidsa
mol wtg
3 h
overnight
polycondensation
in a vacuum
specific
optical
polycondensation
at 180 °C
polymerb
P(LA:RA),
w/w
melting
tempe
t,°C
intrinsic
viscosity,f
dL/g
lactic
acidc
rotation,d
[R]20
Mn
Mw
Mn
Mw
D
99:1
L
L
L
L
L
L
L
L
L
-
-139.5
-135
-131.5
-131
-117.5
-106
-94
98
88
85
80
55
h
0.14
0.18
0.21
0.19
0.18
0.11
0.16
0.13
0.13
0.10
0.18
0.27
900
970
1100
1040
1000
770
950
850
1000
1100
1300
1600
1500
1200
1200
1140
1100
800
1100
950
1200
1700
1600
1900
2000
3100
4200
4100
4000
2500
5800
4200
3500
4300
4400
11 000
2700
4500
6200
5800
5600
3200
7800
5500
4500
5600
5600
13 000
97:3
94:6
90:10
85:15
80:20
70:30
h
60:40
-93
h
50:50
-88
+25
-140
+150
h
100% PRA
100% L-PLA
100% D-PLA
h
L
158
163
D
a RA and LA acids were condensed at 180 °C for 3 h. b w/w ratio of the monomers used for copolymerization. c Lactic acid type. d Optical
rotations of polymers determined by an Optical Activity LTD polarimeter in a 10 mg/mL polymer solution in CHCl3. e Melting temperature
of the copolymers measured by a Fisher apparatus. f Intrinsic viscosity of the polymers in dichloromethane measured in Cannon-Ubbelohde
75 dilution viscometer. g Molecular weight determined by GPC. h Liquid at room temperature.
carbonyl stretching bands. 1H NMR spectra of the
m, C3 -CH2, 2H, m, C13 -CH2, and 3H, d, -CH3, LA), 1.34-
1.25 (16H, m, C4-7 and C14-17), and 0.868 (3H, t, C18 -CH3)
ppm.
polymers fit their composition. The molecular weights,
thermal properties, specific optical rotation of the
polymers, and their intrinsic viscosity are summarized
in Table 1. Copolymerization process did not affect the
optical purity of PLAs. There is a correlation between
intrinsic viscosity and P(LA-RA)s molecular weight and
content. All the liquid polymers had an intrinsic viscos-
ity (in dichloromethane) between 0.10 and 0.16 dL/g
(Table 1). Solid polymers found to be more viscous in
the same conditions. All P(LA:RA)s with up to 20% RA
w/w had an intrinsic viscosity between 0.14 and 0.27
dL/g.
1H NMR (CDCl3, 100% PRA, δ): 5.44-5.30 (2H, m, C9-10,
-CHdCH-), 4.873 (1H, m, C12 HC-O-), 2.309 (2H, t, C2
-CH2), 2.194 (2H, t, C11 -CH2), 2.01 (2H, m, C8 -CH2), 1.603
(2H, m, C3-CH2), 1.446 (2H, m, C13 -CH2), 1.291 (16H, m,
C4-7 and C14-17), and 0.862 (3H, t, C18 -CH3) ppm.
1H NMR (CDCl3, PLA, δ): 5.16-5.15 (1H, q, CH-CH3) and
1.58-1.56 (3H, d, -CH3, LA) ppm.
2.5. Copolymer Synthesis by Transesterification and
Repolymerization by Polycondensation. Transesterifica-
tion of PLA with ricinoleic acid was conducted as follows: 250
mL round-bottomed flask, equipped with a Dean-Stark reflux
condenser and CaCl2 drying tube, was charged with pure
ricinoleic acid and PLA (L-PLA: Mn ) 41 000; Mw ) 91 000)
in the desired ratios (w/w total amount of both compounds was
10 g) and 100 mL of toluene. The ingredients were dried
overnight by refluxing toluene to remove water traces, and
then toluene was removed and in bulk transesterification was
Polymerization of pure RA by thermal polycondensa-
1
tion was confirmed by H NMR (Figure 1). The proton
on C12 is relocated to lower field as a result of ester
bond formation. It was not the only change. Double bond
protons were slightly shifted to upper field upon the
polymerization. It can be explained by long-range
deshielding/shielding effects. The protons close to elec-
tronegative groups, such as hydroxyl, experience a lower
density of shielding electrons and absorb at lower
frequency (deshielding). On the other hand, the protons
distant from such groups, as next to ester bond that was
created during polymerization, absorb at a higher
frequency (all expressed in relation to tetramethylsi-
lane). This 1H NMR analysis allows the determination
of unreacted monomer left in the polymer sample.
1
proceeded for 12 h at 150 °C, followed by GPC and H NMR
analysis. The reaction was stopped as soon as the product
achieved minimal constant molecular weight. Repolymeriza-
tion was carried out by thermal polycondensation. The reaction
flask was connected to an oil pump and heated to 150 °C under
a vacuum of 0.3 mmHg for additional 10 h, followed by GPC.
1H NMR (P(LA-RA) 60:40, δ): 5.47-5.29 (2H, m, C9-10,
-CHdCH-), 5.20-5.00 (1H, q, CH-CH3, LA), 4.90-4.87 (1H,
m, C12 HC-O-), 2.38-2.24 (2H, m, C2 -CH2, and 2H, m,
C11 -CH2), 1.99 (2H, m, C8 -CH2), 1.66-1.40 (2H, m, C3
-CH2, 2H, m, C13 -CH2, and 3H, d, -CH3, LA), 1.30-1.24
(16H, m, C4-7 and C14-17), and 0.866 (3H, t, C18 -CH3).
Transesterification. Copolyester Synthesis. Co-
polyesters of lactic and ricinoleic acids at 9:1 to 5:5 w/w
ratios were synthesized by transesterification, followed
by repolymerization by thermal polycondensation, to
yield solid and viscous materials (Scheme 1). The solid
polymers P(PLA:RA) 9:1-80:20 had an off-white to
yellow brown color, depending on RA content, and
melted at temperatures between 93 and 140 °C. The
viscous polymer P(LA:RA) 5:5 was a yellowish viscous
liquid at room temperature.
Polymers with molecular weights in the range 5000-
11000 were obtained. All polymers possess typical IR
absorption at 1748 cm-1 which corresponds to the ester
carbonyl stretching bands. 1H NMR spectra of the
polymers fit their composition. The physical properties
of the polyesters are summarized in Table 2. Copolym-
erization process did not affect the optical purity of
3. Results and Discussion
Thermal Polycondensation. Copolyester Syn-
thesis. Biodegradable copolyesters of lactic and ricino-
leic acids at various w/w ratios from 99:1 to 50:50, were
synthesized by thermal polycondensation to yield solid
and liquid-viscous materials (Scheme. 1). The solid
polymers P(LA:RA) 99:1-85:15 had an off-white color
and melted at temperatures between 55 and 98 °C. All
viscous polymerssP(LA:RA) 8:2 to 5:5swere clear yel-
low liquids at room temperature that became yellow to
brown as the content of RA content in the copolymer
increased.
Polymers with molecular weights in the range 2000
to 11000 were obtained. All polymers possess typical IR
absorption at 1748 cm-1 corresponding to the ester

5548 Slivniak and Domb
Macromolecules, Vol. 38, No. 13, 2005
Figure 1. 1H NMR spectra: (A) pure ricinoleic acid, with peaks at (a) 5.54, (b) 5.40, and (c) 3.63 ppm; (B) PRA, with peaks at
(a) 5.45, (b) 5.31, and (c) 4.87 ppm in chloroform-d1.
Scheme 1. Synthesis of Poly (RA-LA) by (A) Random Condensation of LA and RA Acids or (B) Transesterification of
PLA with RA and Repolyesterification
Table 2. Properties of Polymers Synthesized by Transesterificationa
mol wtf
12 h
transesterification
150 °C
10 h
specific
optical
repolymerization
150 °C, 0.3 mmHg
polymer
P(l-PLA:RA),b
w/w
melting
ranged
t,°C
intrinsic
viscosity,e
dL/g
rotation,c
25
[R]D
Mn
Mw
Mn
Mw
90:10
-130.5
-106
-81
147
142
111
93
g
0.24
0.14
0.15
0.13
0.15
0.10
3200
1800
1500
1300
1200
1400
4700
4000
2600
2200
2000
2000
11 000
5000
5500
5000
5600
4300
14 000
8000
8100
7000
8200
5600
80:20
70:30
60:40
-68
50:50
-38
+25.5
100% PRA
g
a Polymers synthesized by transesterification of L-PLA Mw ) 91 000 with RA at 150 °C and repolymerization. b w/w ratio of the monomers
used for copolymerization. c Optical rotations of polymers determined by an Optical Activity LTD polarimeter in a 10 mg/mL polymer
solution in CHCl3. d Melting temperature of the copolymers measured by Fisher apparatus. e Intrinsic viscosity of the polymers in
dichloromethane measured in Cannon-Ubbelohde 75 dilution viscometer. f Molecular weight determined by GPC. g Liquid at room
temperature.
PLAs. There is a correlation between intrinsic viscosity
and the P(LA-RA)s molecular weight.
Transesterification and Repolymerization Ki-
netics. RA was incorporated in the PLA by a transes-

Macromolecules, Vol. 38, No. 13, 2005
Lactic Acid and Ricinoleic Acid Based Copolyesters 5549
1
Transesterification kinetics was also followed by H
NMR (Figure 4). As we previously showed (Figure 1),
there are detectable difference in the location of double
bond protons of RA monomer and polymerized RA.
During the transesterification reaction, samples were
withdrawn from the reaction flask and analyzed by GPC
and by 1H NMR. From the integration of 1H NMR
spectra from different transesterification stages (3,7 and
12 h) we can calculate the ratio between reacted and
unreacted RA. According to those calculations, 46%
monomer was left after 3 h of transesterification and
only 13% after 7 h, and no traces of monomer were found
after 12 h of transesterification.
GPC was used to follow the transesterification reac-
tion. RA monomer (MW ) 298), is easily detectable by
GPC and also was well separated from other detectable
molecular weights. The peak area ratio of monomer and
copolymer fit their relative quantities during the trans-
esterification reaction. From this ratio, the percentage
of reacted RA monomer was calculated. The data
calculated from GPC were compared to data that were
Figure 2. Monitoring the molecular weight by GPC during
the transesterification as a function of the RA content. The
data are compared to the calculated molecular weight. Poly-
mers were obtained from the reaction between PLA (Mw
)
91 000) and RA (99% pure) at 150 °C for 12 h. The PLA-RA
ratio is weight to weight.
1
obtained from H NMR (Figure 5), and a good correla-
tion was found.
terification reaction (Scheme 1).The hydroxyl group of
RA attacked ester group of L-PLA (Mn ) 41 000, Mw )
91 000), to form new ester bonds. As the result of
transesterification, RA (B) terminated PLA (A) blocks
were created (AAAABsfirst 12 hstransesterification
reaction). The reaction between RA and PLA was
continued until a constant molecular weight close to the
calculated value was obtained (Figure 2). The calculated
molecular weight assumed that pure ricinoleic acid was
reacted with PLA of an Mw ) 91 000.
Figure 3 shows the change in the molecular weight
of P(PLA-RA) during transesterification and repoly-
merization processes. The molecular weight after 12 h
of transesterification fit the expected molecular weight
reduction. At this stage, the system was connected to a
high vacuum oil pump and repolymerization process
started. The repolymerization condensation was con-
tinued until no additional increase in polymer’s molec-
ular weight was observed. It was found that for all PLA:
RA ratio AAAAB blocks were connected during repolym-
erization reaction. The molecular weights and melting
temperatures of the P(PLA-RA) are shown in Table 2.
Determination of Polymer Structure. Two series
of P(LA-RA)s were synthesized by two different con-
cepts. As it was summarized in Tables 1 and 2, copoly-
esters synthesized by polycondensation and by trans-
esterification resulted in different physical properties
for the corresponding copolymers.
These physical differences indicate different polymer
structure: comonomer sequences and degrees of crystal-
linity. Those two parameters were determined using 1H
NMR and DSC. Whereas NMR spectroscopy could
identify statistical segments in the polymeric chain,
DSC was used to determine the thermal properties and
degree of crystallinity.24
1
Copolymer composition was verified by H NMR by
the integration ratio of the peaks at 5.15 ppm (one
proton of PLA) and the peaks at 4.87 ppm one proton
(C12) of PRA. The copolymer composition in the poly-
mers was identical with the comonomer ratio in the
weighted amounts. The relative degree of crystallinity
of the copolymers depends directly on PLA block size -
degree of randomness. To correlate the NMR spectra
Figure 3. Molecular weight change during the P(PLA-RA)s preparation by transesterification. The reaction was monitored by
GPC during the transesterification and repolymerization reactions between PLA and RA as a function of time (see Scheme 1).
Area A shows transesterification, while area B shows repolymerization.

5550 Slivniak and Domb
Macromolecules, Vol. 38, No. 13, 2005
Figure 4. 1H NMR spectra: (a) pure ricinoleic acid monomer; (b) 3 h of transesterification reaction, PLA-RA 60:40 w/w, 46%
monomer left; (c) 7 h of transesterification reaction, PLA-RA 60:40 w/w, 13% monomer left; (d) 12 h of transesterification reaction,
1
PLA-RA 60:40 w/w, no monomer left. All spectra performed in H NMR 500 MHz with chloroform-d1 as solvent. Calculations
performed according to peak integration.
Table 3. Structure of NMR Checked Comonomer Units
Figure 5. Transesterification reaction progress as a function
of the reacted RA content and reaction time as monitored by
1
GPC and H NMR.
the copolymers’ NMR spectra are attributed to long-
range deshielding/shielding interactions. The LA-LA
interactions are known from the pure PLA’s NMR
spectra. Ester bonds on both sides (LA-LA) cause the
deshielding effect and the proton absorbs at a lower
frequency. If we compare LA-RA and RA-LA, in the
case of (AAAB) LA-RA, the R-proton of LA is located
closer to the RA double bond and experiences a lower
density of shielding electrons and absorbs at lower
frequency (deshielding). For (BAAA) RA-LA, the R-pro-
ton distant from such groups, such as that next to the
-CH2- chain, absorbs at a higher frequency (all ex-
pressed in relation to tetramethylsilane).
to the composition and the frequency of occurrence of
specific comonomer sequences and to determine the
degree of randomness and the number-average sequence
mathematical models was used. If the copolymer is not
strictly alternating or blocklike, a randomly selected
pair of comonomer units in the polymer chain may be
represented as follows: RA-RA, RA-LA, LA-RA, LA-
LA. To simplify the calculations, only the last three
options were considered. The structures of those comono-
mer units are summarized in Table 3.
An examination of the 1H NMR spectra of P(LA-RA)s
(Figure 6) revealed three well separated quartets at
5.17, 5.10, and 5.05 ppm (one R-proton of PLA). The
separation of the PRA proton peaks (4.87 ppm) was less
pronounced. The PLA homopolymer has only one quar-
tet at 5.15 ppm (Figure 6a). These additional peaks in
As the w/w ratio and preparation method of the
copolymers varied, the integration ratios of these peaks
changed. Figure 6 represents the relevant peaks for
P(LA:RA)s 60:40 w/w, synthesized by condensation and
1
by transesterification. The integration data of the H

Macromolecules, Vol. 38, No. 13, 2005
Lactic Acid and Ricinoleic Acid Based Copolyesters 5551
Figure 6. 1H NMR spectra: (a) PLA; (b) PRA; (c) P(LA-RA) 60:40 w/w synthesized by thermal polycondensation; (d) P(LA-RA)
60:40 w/w synthesized by transesterification followed by condensation; (e) P(LA-RA) 60:40 w/w synthesized by ROP. Spectra
performed on a Varian NMR 500 MHz in chloroform-d1. Calculations performed using peak integration.
NMR spectra of P(LA-RA)s 90:10 to 50:50 w/w, syn-
thesized by both methods, were calculated in the fol-
lowing manner.
II. Number-Average Sequence Length of a Mono-
mer Sequence (Ln)
1. Unconditional Probability. [p(LA) and p(RA)]
are the probabilities of a randomly selected monomer
unit in the polymeric chain to be either LA or RA. This
could be determined by the overall integration ratio of
LA to RA.
LLA ) 1/p(RA-LA)
(3)
For the P(LA-RA)s synthesized by thermal polycon-
densation and by transesterification with w/w ratios
from 9:1 to 5:5, the results are presented in Table 4.
Obtained results can be summarized in the following
way. Polymers prepared by thermal polycondensation
have alternating tendency (h > 1), [AABABABBAAA-
BABB]. Polymers prepared by transesterification obtain
block character (h < 1), [AAABAAABAAABB] apart
from P(PLA:RA) 50:50 w/w. Its alternating tendency is
a result of the relatively high RA content. The number-
average sequence length (LLA) of lactic acid decreased
from 12 to 4 for P(LA-RA)s from 9:1 to 5:5, respectively.
The number-average sequence length (LLA) of lactic acid
decreased from 50 to 17 for P(LA-RA)s prepared by
transesterification from 9:1 to 5:5, respectively. Se-
quence length is an additional evidence to random and
block character of the polymers prepared by random
polymerization vs transesterification, respectively.
The same calculations were performed on previously
studied RA based copolyesters synthesized by ring-
opening polymerization (ROP).21 The results are given
in Table 4. RA based polyesters synthesized by ROP had
a block character (h < 1). Actually they obtain a
p(LA) + p(RA) ) 1
(1)
2. Probability. [p(LA-LA), p(LA-RA), and p(RA-
LA)] are the probabilities that the randomly selected
connection was either (LA-LA), (LA-RA), or (RA-LA).
This probability could be determined from the integra-
tion ratios of the relevant peaks, as presented in Figure
6.
From these probabilities and the feed ratios it is
possible to calculate the following:
I. Degree of Randomness (h) of the Formed
Polymer
h ) p(LA-RA)/p(LA)p(RA)
(2)
h < 1 means block character of the copolymer; h ) 1
means the polymer takes a random distribution; h > 1
means alternating tendency; and h ) 2 means a full
alternating copolymer.

5552 Slivniak and Domb
Macromolecules, Vol. 38, No. 13, 2005
Table 4. Comonomer Sequence Distribution of the P(LA-RA)sa
feed ratio
of LA-RA
in the
probability
probability
of finding
the [LA-RA]
unit,
probability
of finding
the [RA-LA
unit,
of finding
the [LA-LA]
unit,
av
block
length,d
LLA
degree of
randomness,c
h
polymer,b
p(LA)
p(LA-LA)
p(LA-RA)
p(RA-LA)
Condensation
0.9
0.8
0.7
0.6
0.5
0.80
0.70
0.53
0.48
0.40
0.12
0.17
0.36
0.40
0.35
0.08
0.13
0.11
0.12
0.25
12.5
7.7
9
8.3
4
1.33
1.06
1.71
1.67
1.4
Transesterificatione
0.9
0.8
0.7
0.6
0.5
0.90
0.82
0.77
0.76
0.64
0.08
0.14
0.20
0.20
0.30
0.02
0.04
0.03
0.04
0.06
50
0.89
0.87
0.95
0.83
1.2
25
33.3
25
16.7
Ring Opening Polymerization (ROP)
0.9
0.8
0.7
0.6
0.5
0.99
0.96
0.97
0.92
0.94
0.01
0.04
0.03
0.08
0.06
0
0
0
0
0
f
f
f
f
f
0.11
0.25
0.14
0.33
0.24
a Probability, degree of randomness, and LA average block length were calculated based on LA:RA (w/w) feed ratios (9:1 to 5:5), according
to eqs 1, 2, and 3 as presented in the text.23 It was assumed that ricinoleic acid is pure and PLA molecular weight Mw ) 91 000. b Equation
1. c Equation 2. d Equation 3. e PLA. f Limited only by molecular weight; polymer with AAABBB structure.
Figure 7. Thermal properties of P(LA-RA)s as determined by DSC. Column 1. P(LA-RA)s synthesized by transesterification
followed by repolymerization by condensation. Column 2. P(LA-RA)s synthesized by direct polycondensation of RA + LA acids,
and Column 3: P(LA-RA)s synthesized by ring-opening polymerization (ROP).21 The P(LA-RA) copolymer compositions were as
follows: (a) 100:0; (b) 90:10; (c) 80:20; (d) 70:30; (e) 60:40; (f) 50:50 w/w of LA:RA.
[AAAAABB] diblock structure. Such a structure is the
result of the low reactivity of the sterically hindered RA
line brushlike domains along the polymer chain, thus
increasing the RA content decreased the melting point
and the crystallinity of the polymers. Figure 7 shows
that polymers (Figure 7(1,3)) possessing a long enough
LA block form detectable crystalline domains in the
polymer where copolymers (Figure 7(2c-f)), with 20%
and higher RA content do not have LA blocks long
enough to form a detectable crystalline area in the
polymer.
1
lactone. H NMR analysis of those polyesters confirms
the diblock structure (Figure 6e). No [RA-LA] connec-
tions were found. The polyester contained only a small
amount of [LA-RA] connections.
DSC revealed a crystalline structure for polyesters
synthesized by transesterification and by ROP (Figure
7). For polyesters synthesized by random condensation
only P(LA-RA) 90:10 w/w contained crystalline areas.
This information is correlated with the 1H NMR analy-
sis where polymers containing relatively long LA blocks
obtain crystalline areas. In P(LA-RA)s, only LA blocks
are able to crystallize. Ricinoleic acid structure is
sterically hindered where RA blocks formed noncrystal-
1. P(LA-RA)s are synthesized by transesterification
followed by repolymerization by condensation.
2. P(LA-RA)s are synthesized by direct polyconden-
sation of RA + LA acids.
3. P(LA-RA)s are synthesized by ring-opening po-
lymerization (ROP).21

Macromolecules, Vol. 38, No. 13, 2005
Lactic Acid and Ricinoleic Acid Based Copolyesters 5553
(6) Kumar, N.; Ravikumar, M. N. V.; Domb, A. J. Adv. Drug
Deliv. Rev. 2001, 53, 23-44.
The copolymer compositions were as follows: (a) 100%
PLA; (b) P(LA-RA) 90:10 w/w; (c) P(LA-RA) 80:20 w/w;
(d) P(LA-RA) 70:30 w/w; (e) P(LA-RA) 60:40 w/w; (f)
P(LA-RA) 50:50 w/w.
(7) Penczek, S.; Szymanski, R.; Duda, A.; Baran, J. Macromol.
Symp. 2003, 201, 261-269.
(8) Seppala, J.; Helminen, A.; Korhonen, H.,. Macromol. Biosci.
2004, 4, 208-217.
Conclusions
(9) Sodergard, A.; Stolt, M. Prog. Polym. Sci. 2002, 27, 1123-
1163.
Copolymers of lactic acid (LA) and ricinoleic acid (RA)
were synthesized by random condensation of RA and
LA or by transesterification and repolymerization to
yield random and multiblock copolymers, respectively.
Copolymers prepared by ring opening copolymerization
reported recently resulted in an AB diblock LA-RA
copolymer.21 The copolymers of similar RA-LA compo-
sition and molecular weights showed significant differ-
ences in thermal properties, where viscous liquids at
room temperature with RA content as low as 20%. These
polymers, particularly the liquid polymers, may have
use as sealants and as injectable carriers of drugs.
(10) Hiltunen, K.; Harkonen, M.; Seppala, J. V.; Vaananen, T.
Macromolecules 1996, 29, 8677-8682.
(11) Hiltunen, K.; Seppala, J. V.; Harkonen, M. J. Appl. Polym.
Sci. 1997, 64, 865-873.
(12) Tuominen, J.; Seppala, J. V. Macromolecules 2000, 33, 3530-
3535.
(13) Grijma, D.; Nijenhuis, A.; Pennings, A. Polymer 1990, 31,
2201-2206.
(14) Grijma, D.; Pennings, A. Polym. Bull. 1991, 25, 335-341.
(15) Grijma, D.; Pennings, A. Macromol. Chem. Phys. 1994, 195,
1633-1647.
(16) Buchholz, B. J. Mater. Sci. 1993, 4, 381-388.
(17) Teomim, D.; Domb, A. J. Biomacromolecules 2001, 2, 37-
44.
Acknowledgment. The authors thank Drs. Gibson
and Nagegra for their help on the NMR analysis. This
work was supported in part by a grant from the BSF,
US-Israel binational foundation.
(18) Teomim, D.; Nyska, A.; Domb, A. J. J. Biomed. Mater. Res.
1999, 45, 258-267.
(19) Borsotti, G.; Guglielmetti, G.; Spera, S.; Battistel, E. Tetra-
hedron 2001, 57, 10219-10227.
(20) Krasko, M. Y.; Shikanov, A.; Ezra, A.; Domb, A. J. J. Polym.
Sci., Polym. Chem. 2003, 41, 1059-1069.
References and Notes
(21) Slivniak, R.; Domb, A. J. Biomacromolecules 2005, 6, 1679-
1688.
(1) Domb, A. J.; Amselem, S.; Maniar, M. Polym. Biomater. 1994,
399-433.
(22) Leong, K. W.; Brott, B. C.; Langer, R. J. Biomed. Mater. Res.
1985, 19, 941-955.
(2) Domb, A. J.; Nudelman, R. J. Polym. Sci., Polym. Chem. 1995,
33, 717-725.
(23) Ron, E.; Mathiowitz, E.; Mathiowitz, G.; Domb, A.; Langer,
(3) Teomim, D.; Domb, A. J. J. Polym. Sci., Polym. Chem. 1999,
37, 3337-3344.
R. Macromolecules 1991, 24, 2278-2282.
(24) Bovey, F. A. Chain Structure and Conformation of Macro-
(4) Duda, A.; Penczek, S. Polimery 2003, 48, 16-27.
(5) Ikada, Y.; Tsuji, H. Macromol. Rapid Commun. 2000, 21,
117-132.
molecules; Academic Press: New York, 1982.
MA0503918