Solvent-free lipase-mediated synthesis of six-membered cyclic carbonates from trimethylolpropane and dialkyl carbonates

Article abstract of DOI:10.1039/c0gc00783h

Six-membered cyclic carbonates with hydroxyl and/or alkoxycarbonyloxy groups, as potential monomers for polyurethanes and polycarbonates, were prepared by the reaction between trimethylolpropane (TMP) with dimethyl carbonate (DMC) or diethylcarbonate (DEC) mediated by immobilized Candida antarctica lipase B, Novozym 435 in a solvent-free medium. The dialkyl carbonate served as a solvent for the reaction. The solubility of TMP at 50°C was 565 mg mL^-1 in DMC, and 64.8 mg mL^-1 in DEC. Reactions using biocatalyst concentrations of 10, 20 and 40% (w/w of TMP) showed similar profiles, with a linear increase in the conversion of TMP to 90% within 24 h, and complete conversion within 48 h. At biocatalyst concentrations of 2.5 and 5% (w/w of TMP), the main products were mono-carbonated TMP (3) and/or cyclic carbonate (4), while di and tri-carbonated TMP (5, 7) were obtained at lipase concentrations of 20 and 40%. The reactivity of DEC was lower than that of DMC, but led to higher selectivity in production of 3 or 4. A large fraction of the linear carbonates in the product mixture were cyclized by disproportionation involving heating at 60-80°C without any catalyst. The total yield of cyclic carbonates was about 85% after thermal treatment at 80°C. This process, consisting of lipase-catalyzed transesterification and thermal disproportionation, provides a novel and more environmentally friendly approach for the synthesis of cyclic carbonates without using toxic organic solvents, phosgene or isocyanate. The Royal Society of Chemistry.

Full text of DOI:10.1039/c0gc00783h

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 976
www.rsc.org/greenchem
PAPER
Solvent-free lipase-mediated synthesis of six-membered cyclic carbonates
from trimethylolpropane and dialkyl carbonates†
Sang-Hyun Pyo,*a Per Persson,b Stefan Lundmarkb and Rajni Hatti-Kaula
Received 7th November 2010, Accepted 11th January 2011
DOI: 10.1039/c0gc00783h
Six-membered cyclic carbonates with hydroxyl and/or alkoxycarbonyloxy groups, as potential
monomers for polyurethanes and polycarbonates, were prepared by the reaction between
trimethylolpropane (TMP) with dimethyl carbonate (DMC) or diethylcarbonate (DEC) mediated
R
by immobilized Candida antarctica lipase B, Novozymꢀ 435 in a solvent-free medium. The dialkyl
carbonate served as a solvent for the reaction. The solubility of TMP at 50 ◦C was 565 mg mL-1 in
DMC, and 64.8 mg mL-1 in DEC. Reactions using biocatalyst concentrations of 10, 20 and 40%
(w/w of TMP) showed similar profiles, with a linear increase in the conversion of TMP to 90%
within 24 h, and complete conversion within 48 h. At biocatalyst concentrations of 2.5 and 5%
(w/w of TMP), the main products were mono-carbonated TMP (3) and/or cyclic carbonate (4),
while di and tri-carbonated TMP (5, 7) were obtained at lipase concentrations of 20 and 40%. The
reactivity of DEC was lower than that of DMC, but led to higher selectivity in production of 3 or
4. A large fraction of the linear carbonates in the product mixture were cyclized by
disproportionation involving heating at 60–80 ◦C without any catalyst. The total yield of cyclic
carbonates was about 85% after thermal treatment at 80 ◦C. This process, consisting of
lipase-catalyzed transesterification and thermal disproportionation, provides a novel and more
environmentally friendly approach for the synthesis of cyclic carbonates without using toxic
organic solvents, phosgene or isocyanate.
elastomers exhibiting improved hydrolytic stability have been
Introduction
developed. Due to their toughness, durability, biocompatibility,
The synthesis of carbonate analogues, especially five- and six-
and improved biostability, they have been incorporated into a
membered cyclic carbonates, has received much attention lately
wide variety of implantable biomedical devices.10 A demand has
due to their potential application as monomers for environmen-
now emerged for new types of polyurethanes providing unaltered
tally benign production of polycarbonates and polyurethanes.1,2
properties, produced without the use of toxic starting materials.
These types of polymers are mainly produced using toxic
phosgene and/or isocyanates.3–5
Aliphatic polycarbonates are tough, dimensionally stable
Among the alternative greener routes that have been pro-
posed, a popular way to synthesize polycarbonate is the ring-
opening polymerization (ROP) of cyclic carbonates in bulk
thermoplastics widely used in engineering and optical appli-
or solution, usually using metallic compounds as catalysts.11
cations. Aliphatic polycarbonates and their copolymers are
Use of metal-free or low-toxicity catalyst/initiators systems for
biodegradable and recyclable,6,7 and are expected to find use in
the ROP of carbonates, such as alcohols/diols, and acid, have
the biomedical field because of their biocompatibility and low
increasingly attracted attention. The ROP initiated by alcohols
toxicity.8,9 Polyurethanes are widely used in a variety of applica-
in the absence of a catalyst has been extensively investigated,
tions such as foams, seals, and high-performance coatings and
since the use of hydroxyl groups (OH) in the alcohols as initiators
adhesives. In recent years, a variety of biomedical polyurethane
allows for control of the molar mass, improvement in the
hydrophilicity/degradability, and further functionalization of
the resulting biodegradable polyesters and polycarbonate.11
aDepartment of Biotechnology, Lund University, Box 124, SE-221 00,
Lund, Sweden. E-mail: Sang-Hyun.Pyo@biotek.lu.se; Fax: +46
46-222-4713; Tel: +46 46-222-8157
Lately, a number of reports have appeared on the syn-
thesis of five-membered and six-membered cyclic carbonates
by a phosgene-free route.12–17 For use in the ROP process,
however, six-membered cyclic carbonates are preferred to five-
membered carbonates because they are less thermodynamically
bPerstorp AB, 284 80, Perstorp, Sweden
† Electronic supplementary information (ESI) available: Tables S1–S4;
Fig. S1. See DOI: 10.1039/c0gc00783h
976 | Green Chem., 2011, 13, 976–982
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
stable than their ring-opened polymers, and thus retain CO2
during the polymerization process.2 Synthesis of six-membered
trimethylene carbonate is traditionally achieved by reacting 1,3-
propanediol with phosgene or its derivatives. Among the other
reactions studied, metal-catalysed coupling of oxetanes such as
trimethylene oxide with carbon dioxide has given high yields of
trimethylene carbonate.2 Endo et al.18 have reported a method
of cyclic carbonate synthesis from propane-1,3-diols and ethyl
chloroformate in the presence of a stoichiometric amount
of triethylamine. Transesterification of propane-1,3-diols with
dialkyl carbonate catalyzed by metal or organo-catalysts has
been proposed as a more environmentally benign procedure.19,20
The lipase-catalysed transesterification reaction between dialkyl
carbonate and 1,3-diol in an acetonitrile–toluene solvent system
for the synthesis of a cyclic trimethylene carbonate (1,3-dioxane-
2-one) monomer with/without a methyl substituent has also
been reported using very high concentration of the biocatalyst
(900% w/w of the diol).21
Fig. 1 Solubility of TMP in DMC (᭜) and DEC (᭡) with increase
in temperature. Experimental details are stated in the text. TMP was
soluble in DMC at 50 ◦C up to a concentration of 565 mg mL-1.
However, the syntheses of six-membered cyclic carbon-
ates with functional groups from poly-functional alcohols
such as trimethylolpropane (TMP) or pentaerythritol (PE)
have required more complicated methods with low yields.22,23
Polycyclic six-membered carbonates could be prepared by
radical polymerization of acrylic monomers with pendant
cyclic carbonate groups.22 As a different approach, tris- or
tetrakis(alkoxycarbonyloxy) derivatives, obtained from catalytic
transesterification of the polyol, TMP and diethylcarbonate
(DEC), have been subjected to thermal disproportionation using
Aerosil 200 at 200–220 ◦C followed by distillative depolymerisa-
tion under reduced pressure, to give the cyclic product 5-ethyl-5-
ethoxycarbonyloxymethyl-1,3-dioxan-2-one with a low yield.23
This paper presents a study showing the possibility of the
synthesis of six-membered cyclic carbonates with functional
groups by lipase-mediated reaction between trimethylolpropane
(TMP) with dimethyl carbonate (DMC) or DEC in a solvent-
free medium followed by thermal disproportionation without
any catalyst.
to 1 mL DMC), and in DEC up to 64.8 mg mL-1. At 60 ◦C
there was a phase change of TMP (melting point 57 ◦C), which
was completely miscible with DMC (2 g TMP added to 1 mL
DMC) but formed a biphasic system with DEC. The polyol
concentration in the upper DEC phase was 138.3 mg mL-1.
The solubility of TMP was thus sufficient for reaction at 50–
60 ◦C, which is an optimum temperature range for the activity
of immobilized lipase.26
The stability of a biocatalyst is usually higher in more hy-
drophobic solvents (i.e. those with higher logP – the logarithm of
the partition coefficient of solvent between water and octanol),
such as toluene (2.5) and n-hexane (3.5). In contrast, solvents
with lower logP values such as pyridine, dimethyl formamide
(DMF, -1.0), dimethyl sulfoxide (DMSO, -1.3) and acetonitrile
(-0.33) can solubilise many polar molecules like TMP, but often
inactivate the enzyme by their ability to remove water molecules
of hydration and promote accumulation of water in the reaction
medium.27–29 Hence, the balance between hydrophobicity and
solubility needs to be considered in enzymatic reactions. The
logP values of DMC and DEC were calculated as 0.23 and
1.22, respectively, according to Parham’s method.30 A stability
Results and discussion
R
test of immobilized Candida antarctica lipase B, Novozymꢀ 435
(N435), in DMC, DEC, acetonitrile and n-hexane at 60 ◦C
showed the residual (esterification) activity after 72 h incubation
to be 67.6, 77.6, 63.1 and 82.4%, respectively. The biocatalyst was
more stable in DEC than in DMC and acetonitrile. As expected,
the highest stability was obtained in n-hexane, a solvent that
did not dissolve TMP and provided poor conditions for the
enzymatic reaction.
Solubility/miscibility of TMP in DMC and DEC
Solvents such as THF, acetonitrile and toluene have been
employed to enhance co-solubilization of substrates in biocat-
alytic syntheses of cyclic carbonates.15,19 Although some envi-
ronmentally benign alternatives to the use of organic solvents in
organic syntheses, e.g. ionic liquids, are proposed for enzymatic
reactions,24 the most desirable strategy from environmental and
economic perspectives is to carry out the reaction without a
solvent. In the present system, the dialkyl carbonate can serve
as a solvent due to a relatively low boiling point (DMC: 90 ◦C
and DEC: 126 ◦C) allowing easy removal and recycling. The
solubility of TMP in DMC and DEC was thus investigated at
different temperatures. DMC is an inexpensive, environmentally
benign chemical having interesting solvating properties, low
toxicity and high biodegradability.25
Structure elucidation of the products and proposed reaction
pathway
Due to the unavailability of standard compounds, TMP was
initially reacted with excess of DMC or DEC at 60 ◦C using N435
as catalyst, to determine the identity of the products formed. All
the reaction components were analysed by gas chromatography,
and identified by GC-MS and 1H-NMR (Table S1†).
As seen in Fig. 1, the TMP solubility increased at temperatures
GC analysis showed the presence of several peaks (Fig. 2A).
All product peaks on GC chromatograms were purified by silica
flash chromatography, and the chemical shifts of these products
◦
hig◦her than 30 C, and the TMP was fully soluble in DMC at
50 C at concentrations up to 565 mg mL-1 (1.3 g TMP added
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 976–982 | 977
©

View Article Online
Fig. 2 Gas chromatogram (A) and scheme of the possible reaction pathway (B) for lipase-catalyzed reaction of TMP and dialkyl carbonate. R =
CH3 (DMC) or CH2CH3 (DEC).
were elucidated by 1H-NMR (Table S1), which agreed with the
structures shown in Fig. 2B. A possible reaction pathway was
thus proposed. A lipase-catalysed transesterification reaction
between the TMP (1) hydroxyl groups and the dialkyl carbonate
(2) would result in the formation of linear carbonates, which
appeared as products 3 and 5 (TMP mono- and dicarbonate,
respectively). The transesterification products were further con-
verted to the corresponding cyclic carbonates 4 and 6 (Fig. 2). 7
(TMP-tricarbonate) was formed in extremely small amounts.
In contrast to the GC results, GC–MS data did not reveal any
masses corresponding to the linear carbonates, but confirmed
the presence of cyclic carbonates. This implies that the GC–MS
Fig. 3 Effect of NovozymꢀR 435 concentration on conversion of TMP
data of the linear carbonates, 3, 5 and 7, gave masses of demethy-
lated and decarbonated products, 160.9 (M - 32 (-HOCH3)),
219.0 (M - 32(-HOCH3)) and 233.0 (M - 76 (-HOCOOCH3)),
respectively. This was probably due to exposure of the product
samples to the high temperature in the injection port. Masses of
160.9 and 219.0 correspond to that of cyclic carbonates, 4 and
6, respectively, obtained from reaction of TMP with DMC and
DEC, respectively (Table S1†).
during reaction with DMC at 60 ◦C. The biocatalyst was used at 2.5%
(ꢀ), 5% (᭜), 10% (᭡), 20% ( ) and 40% ( ) concentration (w/w),
᭹
respectively, with respect to TMP.
10% (w/w) biocatalyst, and 100% for 20 and 40% biocatalyst)
within 48 h. Conversions of 98.9% and 83.3% were achieved
with 5% (w/w) and 2.5% N435, respectively, at 120 h. Excellent
reproducibility of the results was demonstrated.
The proportions of products formed varied depending on the
biocatalyst concentration and reaction time (Table S2†). In the
reaction with 2.5 and 5% N435, the TMP-monocarbonate 3 was
the predominant product (62%) along with other products in
the order 4 > 5 > 6 until 48 h reaction time, and subsequently
the amount of 3 decreased due to conversion to the above
products. Cyclic carbonate formation seemed to be preferred
rather than further transesterification of 3. Compound 7 (TMP-
tricarbonate) was not detected except in the sample from 120 h
R
Effect of Novozymꢀ 435 concentration on the reaction between
TMP and DMC
N435 was used at concentrations of 2.5, 5, 10, 20 and 40%
(w/w of TMP) for reaction with DMC at 60 ◦C. As shown
in Fig. 3, reactions employing higher biocatalyst amounts (10,
20 and 40%) showed nearly identical time courses, with a linear
increase of the conversion of TMP to reach 90% within 24 h, and
continued reaction leading to complete conversion (98.5% for
978 | Green Chem., 2011, 13, 976–982
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
reaction with 5% N435. With 10% N435, concentrations of
the products formed at 48 h varied in the order 5 > 3 > 6
> 4 > 7, indicating increased activity of the biocatalyst for
transesterification (Table S2†). With continued incubation, the
levels of 5, 6 and 7 increased, while that of 3 decreased (Table S2,†
Fig. 4). With further increase in the biocatalyst concentration
(20, 40%), the main product was TMP-dicarbonate 5, along
with all the other products. 7 was formed in significant amounts,
and levels of both 5 and 7 increased when the biocatalyst
concentration was increased from 20 to 40%, indicating that the
transesterification activity was dominating over the subsequent
cyclic carbonate formation. Among the cyclic carbonates, the
levels of 6 were higher than that of 4 due to higher proportion
of 5 formed.
6 being greater than 4. Meanwhile, cyclic carbonates 4 and 6 were
obtained as the main products (about 75% of total) during the
reaction at 70 ◦C. This was an indication that high temperature
is important for decarbonation of the linear carbonates to cyclic
carbonates. Hence, the optimum temperatu◦re is 50 ◦C to produce
linear mono-carbonated product, and 70 C to produce cyclic
carbonates. It may be possible to utilize the entire product
mixture containing high proportion of cyclic carbonates for a
coating formulation, but when there is a need for high-purity
cyclic carbonates the components of the product mixture can
most probably be separated by distillation or crystallization.
R
Reaction between TMP and DEC catalysed by Novozymꢀ 435
Although both DMC and DEC can undergo transesterification
with TMP, they have different physicochemical properties. DEC
is more hydrophobic, larger in size, and dissolves TMP to a
lower extent. As shown in Fig. 5, DEC had lower reactivity than
DMC. The conversion of TMP during reaction with DEC using
20% (w/w) N435 reached 98.7% at 120 h, while reaction under
similar conditions with DMC using 5–10% (w/w) lipase reached
same conversion at 48 h. Nevertheless, the trend of product
formation was almost the same as for DMC (Table S4†, Fig. 6).
Fig. 4 Conversion of TMP (᭡) and formation of products, 3 (ꢀ), 4 ( ),
᭹
5 (᭜), 6 ( ), and 7 (ꢁ), respectively, during the reaction of the polyol
with DMC catalysed by 10% (w/w of TMP) NovozymꢀR 435 at 60 ◦C.
The immobilized lipase concentrations used here are much
lower than that used in the earlier reports of the synthesis
of six-membered trimethylene carbonate (1,3-dioxane-2-one)
monomer (900% w/w of 1,3-diol)21 and five-membered glycerol
carbonate (55% w/w of glycerol).15 In the former case, oligomers
were formed as side-products, which were recovered for further
transformation to cyclic carbonate, obtained in a final yield of
53%.21
Fig. 5 Comparison of the reaction profiles of TMP with DEC using
2.5% (ꢀ), 5% (᭜), 10% ( ), and 20% (w/w of TMP) ( ), respectively,
᭹
of NovozymꢀR 435 with the reaction between TMP and DMC using 10%
N435 (᭡) at 60 ◦C.
Effect of reaction temperature on reaction yield and product
selectivity
The optimum temperature for the activity of N435 is reported
to be around 70–80 ◦C, although lower temperatures are
recommended due to the risk of thermal inactivation.26 However,
the optimum reaction rate varies depending on the kind of
reaction, properties of substrates and solvents used. The reaction
efficiency and product profile were investigated for the reaction
of TMP with DMC using 10% (w/w) N435 at 50, 60 and
70 ◦C, respectively. The reaction rate was slightly higher at
60 ◦C than at the other temperatures (Fig. S1†). Irrespective
of the temperature, a maximum TMP conversion of 98% was
obtained, although with significantly different proportions of
products (Table S3†). The product profile obtained at 50 ◦C was
in the order 3 > 5 > 4 > 6, while at 60 ◦C, higher concentrations
of 5 compared to 3 were obtained due to higher enzyme activity.
The levels of the cyclic carbonates were still in a lower range, with
Fig. 6 Conversion of TMP ( ) and formation of products, 3¢ (ꢀ), 4
᭹
(᭡), 5¢ (᭜), and 6¢ ( ), respectively, during the reaction of the polyol with
DEC using 10% (w/w of TMP) NovozymꢀR 435 at 60 ◦C. The marks (¢)
indicate that the products were from reaction with DEC (R CH2CH3).
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 976–982 | 979
©

View Article Online
Table 1 Disproportionation of carbonates by thermal treatment
Carbonate products
Cyclic (%)
Linear (%)
Run
Reaction temp. (◦C)
Starting Material
60
70
80
80
Reaction time (h)
TMP
4
6
3
5
7
1a
2
3
4
5
2.6
3.9
4.3
3.9
3.1
35.3
41.4
49.6
55.3
63.6
22.7
23.6
24.5
23.3
21.7
17.1
13.1
10.6
8.2
21.0
16.6
9.4
7.6
4.6
1.3
1.4
1.6
1.7
1.7
48
48
48
144
5.3
a Starting material (Fig. 2A) was prepared from 5 g TMP in 40 mL DMC using 10% NovozymꢀR 435 at 60 ◦C and used for thermal treatment.
Compound 3¢ (mono(ethoxycarbonyloxy)-TMP) was the main
product (>87%) with DEC using 2.5–10% (w/w) N435 until
48 h, which on prolonged reaction was converted mainly to 4
(72–44%). No 7¢ (tris(ethoxycarbonyloxy)-TMP) was detected.
With 20% biocatalyst, besides 3¢ (75%), a significant amount of
5¢ (bis(ethoxycarbonyloxy)-TMP) (21%) was also formed, and
on prolonged incubation more 6¢ (mono(ethoxycarbonyloxy)-
TMP cyclic carbonate) than 4 was formed. Increasing the
temperature to 70 ◦C for the reaction with 10% N435, the extent
of TMP conversion was reduced to 44% at 48 h, perhaps due to
deactivation of the biocatalyst, and the product was composed
of almost equal amounts of 3¢ and 4 (Table S4†). These results
point towards the possibility of achieving higher selectivity for
cyclic carbonate 4, although at the expense of lower substrate
conversion. The selectivity could be related to a lower reaction
rate for the enzyme with DEC, at the same time providing more
time for cyclic carbonate formation than with DMC.
nearly constant (Fig. 7, Table 1). After 48 h at 80 ◦C, the total
cyclic carbonate (4 and 6) yield was about 79% (4 being 55%) and
was increased further to 85% with prolonged incubation. This
reaction was focused on equilibrium shift to cyclic carbonate
4, which has a minimum content of the ester (–CO–OCH3),
and this was achieved by continuous removal of the resulting
methanol by evaporation, which prevented the reverse reaction
from taking place. These results are a significant improvement
over the earlier report on the synthesis of six-membered cyclic
carbonate at lower yields (53%), which used large amounts
(900% w/w of the diol) of the biocatalyst.15,21
Conclusions
Synthesis of six-membered cyclic carbonates with functional
groups was achieved at high yields by a lipase-catalyzed reaction
between TMP and dialkyl carbonate combined with thermal dis-
proportionation. TMP is currently produced industrially from
fossil-based butyraldehyde and formaldehyde,32 while DMC is
produced by oxidative carbonylation of methanol (and DEC by
transesterification of DMC with ethanol).31,32 These substrates
could potentially be produced from bio-based butanol and
methanol, once these chemicals are available in larger quantities.
This biocatalytic reaction, which does not use any solvent, or
other toxic materials such as phosgene and isocyanate that
are generally used for the production of polycarbonate and
polyurethane, constitutes a green process for the synthesis of
cyclic carbonates.
A strategy to be considered for achieving higher productivity
and selectivity would be to reach high yields of 3 (or 5) in a
short time (as in Fig. 4) followed by subjecting the product to
high temperature. Further studies on improving the productivity
of both the lipase-catalysed reaction and disproportionation are
ongoing to make it an economically attractive process. Designing
lipase mutants with higher activity and selectivity with polyol
substrates is also planned.
Disproportionation of carbonates by thermal treatment
As the reaction of TMP with the dialkyl carbonates produced
several products including the desired cyclic carbonate, it
was of interest to study if the linear carbonates present in
the product mixture could be converted to the cyclic forms.
Conversion of 1,1,1-tris(ethoxycarbonyloxymethyl)propane to
the cyclic carbonate, 5-ethyl-5-ethoxycarbonyloxymethyl-1,3-
dioxan-2-one, by thermal disproportionation with Aerosil 200 at
200–220 ◦C followed by distillative depolymerisation under re-
duced pressure, reported earlier, is energy-intensive and not suit-
able for polyfunctional alcohols because of possible crosslinking
during transesterification with carbonic acid esters.23
In the present study, disproportionation by thermal treatment
without any catalyst was tested for the formation of cyclic
carbonates. The starting material for this was the reaction
mixture shown in Fig. 2A, prepared from reaction between 5 g
TMP and 40 mL DMC using 10% (w/w) N435 for 72 h reaction
at 60 ◦C, followed by filtration to remove solids including the
immobilized lipase and molecular sieves, and evaporation. It
contained a higher content of cyclic carbonates than prepara-
tions on a small scale. Disproportionation was performed by
heating with shaking at 60–80 ◦C under atmospheric pressure to
allow free evaporation of the resulting methanol. At the higher
temperature (80 ◦C), a significant decrease in the content of
linear carbonates 3 and 5 with a simultaneous increase in cyclic
carbonate 4 was observed, while the content of 6 was maintained
Experimental
Materials
TMP was a product of Perstorp AB (Sweden). Immobilized
R
lipase B from Candida antarctica (Novozymꢀ 435) was a gener-
ous gift from Novozymes A/S (Bagsvaerd, Denmark). Dimethyl
carbonate (97%), diethyl carbonate (97%), and molecular sieves
980 | Green Chem., 2011, 13, 976–982
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 7 Chromatograms for disproportionation of carbonates by thermal treatment at 80 ◦C at reaction time of 15 h (A), 48 h (B), and 144 h (C).
For reaction time 0 h (starting material), see Fig. 2A.
˚
(4 A) were purchased from Sigma–Aldrich Chemical Co.
Lipase-catalysed reaction between TMP and dialkyl carbonate
(St. Louis, MO, USA). HPLC-grade acetonitrile, ethyl acetate
and dichloromethane were purchased from Merck (Germany).
All chemicals were used without further treatment.
TMP (50 mg, 0.37 mmol) was dissolved in 1.5 mL (17.8 mmol)
DMC in a 5 mL vial by shaking at 700 rpm at the required
temperature on a thermomixer. The reaction was started by
adding 150 mg molecular sieves and N435 and continued for
several days. Aliquots were withdrawn at different time intervals
for analysis of reaction components. Effects of biocatalyst
concentration and reaction temperature on TMP conversion
and product profile were investigated. The reactions in DEC
(1.5 mL, 12.4 mmol) were carried out in the same manner.
The amount of immobilized lipase used is always stated as
weight/weight of TMP.
Determination of TMP solubility in dialkyl carbonate
TMP (0.5 g, 3.7 mmol) was added to 1 mL DMC in a 5 mL
capped vial and shaken for 30 min at 700 rpm and 20 ◦C on
a ThermoMixer (MKR 13, HLC Biotech, Germany).◦After
separation of the residual solid TMP by standing at 20 C for
20 min, a small aliquot of the supernatant was withdrawn for
analysis. Then incubation of the mixture was repeated at 30
and 40 ◦C as above. TMP was freely soluble at 50 ◦C at the
above concentration, hence 1–1.3 g was added to 1 mL DMC
for testing the solubility. In DEC, the solubility measurement
was performed in a similar manner (using 0.5 g TMP per 1 mL
DEC) at 20, 30, 40 and 50 ◦C. At 60 ◦C, separate phases of DEC
and melted TMP were generated.
Quantitative analysis and structure elucidation
Quantitative analyses of reaction components were performed
using gas chromatography (GC, Varian 430-GC, Varian, USA)
equipped with a FactorFour Capillary column, VF-1 ms (Varian,
15 m ¥ 0.25 mm) and a flame ionization detector. The i◦nitial
column oven temperature was increased from 50 to 250 C at
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 976–982 | 981
©

View Article Online
◦
a rate of 20 C min-1. The samples, diluted with acetonitrile
4 An Ullmann Encyclopedia: Industrial Organic chemicals: Starting
Materials and Intermediates, Wiley-VCH: New York, 1999, vol. 2,
1045.
5 D. J. Darensbourg, A. Horn Jr and A. I. Moncada, Green Chem.,
2010, 12, 1376.
6 T. Suyama and Y. Tokiwa, Enzyme Microb. Technol., 1997, 20, 122.
7 T. Suyama, H. Hosoya and Y. Tokiwa, FEMS Microbiol. Lett., 1998,
161, 255.
8 Z. Zhang, R. Kuijer, S. K. Bulstra, D. W. Grijpma and J. Feijen,
Biomaterials, 2006, 27, 1741.
9 Y. Yamamoto, S. Kaihara, K. Toshima and S. Matsumura, Macro-
mol. Biosci., 2009, 9, 968.
10 S. A. Guelcher, Tissue Eng., Part B: Rev., 2008, 14, 3.
11 L. Liao, C. Zhang and S. Gong, Eur. Polym. J., 2007, 43, 4289.
12 G. Rokicki, P. Rakoczy, P. Parzuchowski and M. Sobiecki, Green
Chem., 2005, 7, 529.
at a concentration of 0.1–0.5 mg mL-1, were injected in a split
injection mode of 10% at 275 ◦C. The conversion of TMP and
ratio of products were calculated by comparison of peak areas
on the gas chromatograms.
The molecular masses of products were measured by GC–MS
(Varian 431-GC, Varian 210-MS) equipped with a FactorFour
Capillary column, VF-5 ms (Varian, 30 m ¥ 0.25 mm). The initial
column ove◦n temperature was increased from 50 ◦C to 275 ◦C at
a rate of 15 C min-1. The samples diluted as above were injected
at 275 ◦C.
For structure elucidation, the products were isolated from
the reaction mixtures by flash chromatography. The reaction
mixture was loaded onto a silica column (25 mm i.d. ¥
250 mm), which was equilibrated with dichloromethane–ethyl
acetate (2 : 1, v/v). The elution was performed using 3 : 1
dichloromethane–ethyl acetate. The structures of the purified
13 Y. Patel, J. George, S. M. Pillai and P. Munshi, Green Chem., 2009,
11, 1056.
14 J. Sun, S.-I. Fujita and M. Arai, J. Organomet. Chem., 2005, 690,
3490.
15 S. C. Kim, Y. H. Kim, H. Lee, D. Y. Yoon and B. K. Song, J. Mol.
Catal. B: Enzym., 2007, 49, 75.
16 A. J. George, Y. Patel, S. M. Pillai and P. Munshi, J. Mol. Catal. A:
Chem., 2009, 304, 1.
17 B. C. Vieville, J. W. Yoo, S. Pelet and Z. Mouloungui, Catal. Lett.,
1998, 56, 245.
1
products were then determined by H-NMR using 400 MHz
NMR (Bruker, UltraShield Plus 400, Germany).
R
Determination of residual activity of Novozymꢀ 435
18 J. Matsuo, K. Aoki, F. Sanda and T. Endo, Macromolecules, 1998,
31, 4432.
The residual enzyme activity of N435 after incubation in DMC,
DEC, acetonitrile and n-hexane, respectively, was evaluated by
esterification of caprylic acid with ethanol.33 Ten milligrams of
the biocatalyst was incubated in the solvents up to 72 h at 400
rpm and 60 ◦C on a ThermoMixer. The biocatalyst was washed
using 90% ethanol after removal of the solvent, and then reacted
with 200 mM caprylic acid in ethanol (90%) at 40 ◦C. The initial
esterification rate was measured over 15 min by GC analysis
as reported earlier,33 and compared with the fresh biocatalyst
treated using the same procedure.
19 W. H. Carothers and F. J. Van Natta, J. Am. Chem. Soc., 1930, 52,
314.
20 A.-C. Albertsson and M. J. Sjoeling, J. Macromol. Sci., Part A: Pure
Appl. Chem., 1992, 29, 43.
21 H. Tasaki, K. Toshima and S. Matsumura, Macromol. Biosci., 2003,
3, 436.
22 K. Miyazaki, T. Endo, F. Sanda, O. Moriya, T. Fukushima and J. M.
Antonucci, Polym. Prepr., ACS Div. Polym. Chem., 1997, 38, 165.
23 G. Rokicki, T. Kowalczyk and M. Glinski, Polym. J., 2000, 32, 381.
24 R. A. Sheldon, R. M. Lau, M. J. Dorgedrager, F. van Rantwijk and
K. R. Seddon, Green Chem., 2002, 4, 147.
25 R. Bernini, E. Mincione, M. Barontini, F. Crisante, G. Fabrizi and
A. Gambacorta, Tetrahedron, 2007, 63, 6895.
26 Novozymes A/S, Product Sheet, Biocatalysis/2003-03782-01.pdf.
27 S. H. Pyo and D. G. Hayes, Synthesis of saccharide–fatty acid
ester biosurfactants catalyzed by lipase, Biobased Surfactants and
Detergents: Synthesis, Properties, and Applications, AOCS Press:
Champaign, IL, 2009, p. 323.
28 P. Villeneuve, Biotechnol. Adv., 2007, 25, 515.
29 D. G. Hayes, J. Am. Oil Chem. Soc., 2004, 81, 1077.
30 M. Regel-Rosocka and J. Szymanowski, Chemosphere, 2005, 60,
1151.
Acknowledgements
The work was performed within the research program
Greenchem at Lund University supported by the Swedish
Foundation for Strategic Environmental Research (Mistra).
31 P. Werle, M. Morawietz, S. Lundmark, K. So¨rensen, E. Karvinen
and J. Lehtonen, An Ullmann Encyclopedia of Industrial Chemistry:
Alcohols, Polyhydric: Wiley-VCH: Weinheim, 2008, 1.
References
1 B. Barkakaty, K. Morino, A. Sudo and T. Endo, Green Chem., 2010,
12, 42.
2 G. T. Howard, Int. Biodeterior. Biodegrad., 2002, 49, 245.
3 S. Matsumura, Y. Soeda and K. Toshima, Appl. Microbiol. Biotech-
nol., 2006, 70, 12.
32 P. Tundo, F. Arico, A. E. Rosamilia and S. G. L. Rossi, Green
Chemical Reactions, Springer Science: Heidelberg, 2008, p. 213.
`
33 A. E. V. Hagstrom, M. Nordblad and P. Adlercreutz, Biotechnol.
¨
Bioeng., 2009, 102, 693.
982 | Green Chem., 2011, 13, 976–982
This journal is
The Royal Society of Chemistry 2011
©