J BT,889-892,446 Linko

4/12/97 2:05 pm

Page 889

Biotechnology Techniques, Vol 11, No 11, November 1997, pp. 889892

11111
2
3
4
5
6
7
8
9
10111
1
2
3
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
5
6
7
8
9
50111
1
2
3111

Production of trimethylolpropane esters

of rapeseed oil fatty acids by
immobilized lipase
Y.-Y. Linko,1* T. Tervakangas,1 M. Lms2 and P. Linko1
1
Laboratory of Bioprocess Engineering, Helsinki University of Technology, P.O.Box 6100, FIN-02015 HUT, Finland.
Fax +3589462373, E-mail: yu-yen.linko@hut.fi. 2Raisio Chemicals, P.O.Box 101, FIN-21201, Raisio, Finland

The polyol, trimethylolpropane (2-ethyl-2-hydroxymethyl-1,3-propanediol), and a mixture of rapeseed oil fatty acid
methyl esters were transesterified by immobilized lipases without additional organic solvent. The conversion to the
polyol tri-ester with immobilized Rhizomucor miehei lipase Lipozyme IM 20 was about 75% after 24 h at 58C, 5.3 kPa,
with no added water, and the highest conversion of about 90% was reached in 66 h.

Introduction
The interest in the production of biodegradable, environmentally acceptable esters for biodiesel, lubricants,
solvents, surface active agents, etc., from vegetable oils
by lipase (triacylglycerol acylhydrolase, EC 3.1.1.3) biocatalysis has markedly increased during the last few years
(Linko et al., 1995b; Linko and Seppl, 1996). For example, butyl oleate (Linko and Wu, 1996) may be used as
biodiesel additive, PVC plastisizer, water resisting agent,
and in hydraulic fluids. Rapeseed oil fatty acid esters of
2-ethyl-1-hexanol (Linko et al., 1994) can be employed
to replace conventional organic solvents in a number of
detergent applications such as in car shampoos, and as a
solvent for printing ink. Biodegradable esters (Linko et
al., 1995a) are of interest for example as surgical
implants and agricultural plastic films.
The interest in environmentally acceptable biodegradable lubricating base oils has recently rapidly increased
(Mang, 1994). Biodegradable lubricants were first developed for two-stroke outboard engines in the beginning
of 1980s, with the main base fluid composed of
neopentylpolyol esters of branched chain fatty acids.
Eychenne et al. (1996) have recently reviewed the developments in environmentally friendly lubricating oils
based on neopentylpolyols such as neopentyl glycol,
pentaerythritol, and trimethylolpropane. In the middle
of 1980s, biodegradable chain-saw oils based on natural
esters of rapeseed oil were introduced on the market.
Biodegradable trimethylolpropane esters of fatty acids
from sunflower oil (Bongardt et al., 1996) or rapeseed
oil (Lms, 1995) fatty acids can be used for example
in the production of hydraulic fluids. Further, trimethylolpropane esters have been developed as lubricants for
jet turbine (Cooley and Slovinsky, 1961), motor-car
1997 Chapman & Hall

(Leleu et al., 1977), and gas turbine engines (Carr and

DeGeorge, 1989).
Lipase catalyzed transesterification has been previously
proposed for example for the modification of food fats
and oils (Coleman and Macrae, 1977; Yokozeki et al.,
1982), and production of biodegradable solvents (Linko
et al., 1994) and polymers (Linko and Seppl, 1996).
Osada et al. (1987) and Monot et al. (1990) have demonstrated that also hydrophilic polyols can be esterified by
lipase in the presence of an organic solvent such as din-butyl ether or tetrahydrofuran. However, organic
solvents are undesirable from the point of view of practical applications. Preliminary results have shown,
however, that rapeseed oil based trimethylolpropane
esters can also be synthesized by lipase biocatalysis
without an additional organic solvent (Lms et al.,
1995; Linko et al. 1996). We describe now for the first
time the enzymic transesterification between trimethylolpropane and rapeseed oil fatty acid methyl esters in
high trimethylolpropane tri-ester yields using immobilized lipases.
Materials and methods
Materials
Finnish rapeseed oil was obtained from Raisio Group,
Oil Milling Divion. The average fatty acid composition
of the low erucic acid rape seed oil was: oleic acid 57%,
linoleic acid 22%, linolenic acid 12%, palmitic acid
4%, eicosaenoic acid 2%, stearic acid 1%, erucic acid
< 1%, others 1%. The carriers tested for lipase immobilization were: Amberlite XAD-7 and Amberlite IRA94s (Rohm and Haas, Philadelphia, USA), Celite R-630
(Manville, UK), Dowex 66, MWA-1WGR-2 and XUS
40339.01 (Dow Chemical Company, Midland, USA),
Biotechnology Techniques Vol 11 No 12 1997

889

J BT,889-892,446 Linko

4/12/97 2:05 pm

Page 890

Y.-Y. Linko et al.

11111
2
3
4
5
6
7
8
9
10111
1
2
3
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
5
6
7
8
9
50111
1
2
3111

Duolites ES-561 and ES-762 (neutral adsorption resin)

(Dia-Prosim, Vitry Chauny, France), GCC and GDC
200 (weak alkaline anion exchange resin) (Cultor,
Finland), HPA 25 (strong alkaline anion exchange resin)
(Mitsubishi Kasei, Japan), Kieselgel 60 (Merck, Darmstadt, Germany), WA 30 (weak anion exchange resin)
(Mitsubishi Kasei, Japan) and Whatman DE 52 (W &
R Balston, Maidstone, UK).
Chemicals
Trimethylolpropane,
2-ethyl-2-(hydroxymethyl)-1,3propanediol, was obtained from E. Merck (Darmstadt,
Germany). All other reagents were of analytical grade,
unless otherwise indicated.
Enzymes
Lipase preparations used were from Candida rugosa
(powder, hydrolytic activity 8000 U g1) (Biocatalyst,
Pontypridd, UK) and immobilized lipases from Rhizomucor miehei Lipozyme IM 20 (hydrolytic activity
830 U g1 and Candida antarctica Novozym 435 (hydrolytic activity 7400 U g1) (Novo Nordisk, Bagsvaerd,
Denmark).
Enzyme immobilization
Carriers were first washed with boiling, deionized water.
A suitable quantity of the solid lipase preparation was
mixed for 2 h in 0.05 M sodium phosphate buffer (pH
5.8), filtered, and 60 ml of the enzyme solution was
mixed with 40 g buffered carrier in a 250 ml conical
flask for 6 h at 28C. The mixture was filtered, and the
precipitate was washed six times with 60 ml of deionized water and freeze-dried for 30 h to a dry solids
content of at least 99%.
Synthesis of rapeseed oil methyl ester
Rapeseed oil methyl ester was synthesized chemically as
follows: 264 g rapeseed oil was weighed into a 1000 ml
3-necked flask, equipped with a thermometer, condenser, stirrer and sample adapter, and 34 g methanol
was added under stirring. The reaction mixture was
heated to 60C and 0.5% (w/w) alkaline catalyst was
added. After the reaction was completed in 4 h as determined by TLC, the reaction mixture was washed by
acidic water. Glycerol formed was separated and the
excess alcohol was distilled off. The rapeseed oil methyl
ester (melting range 56 to 59C) content of the product
was 96.6%, as determined by HPLC.
Enzymic synthesis of rapeseed oil
trimethylolpropane ester
Transesterification between trimethylolpropane and
rapeseedoil methyl ester was carried out at a reduced

890

Biotechnology Techniques Vol 11 No 12 1997

pressure (2.0 to 13.3 kPa) in 25 ml round bottomed

flasks equipped with a 20 cm vertical condenser (cooling
water temperature 6C) and a magnetic stirrer typically
as follows: trimethylolpropane (0.607 g) was first
dissolved in 0.7 ml (15%, w/w of total mass of the
substrates) of water, after which rapeseed oil methyl
ester (4.00 g) and either solid lipase preparation (40%
w/w) or immobilized lipase (50% w/w) were added.
Reaction was usually carried out either at 37C or 47C
at 5.3 kPa (40 mmHg) with magnetic stirring at 150
rev min1), and the average trimethylolpropane to rapeseed oil methyl ester molar ratio was either 1:3.5 or
1:4.5. The condenser was flushed with about 2 ml
acetone after which the total sample was extracted
5 times with 4 ml acetone. The biocatalyst residue was
removed by centrifuging (1900 g). The supernatant
containing the product trimethylolpropane tri-esters of
rapeseed oil fatty acids was transferred into a 1.5 ml
Eppendorf tubes and stored at 20C for later analyses.
Analytical
Lipase activity (hydrolytic) was determined as follows.
The lipase sample, 2.5 ml of deionized water, and
1.0 ml of McIlvane buffer of pH 7.0 were kept at 37C
for 5 min in a conical flask equipped with a magnetic
stirrer, after which a mixture of 3.0 ml of olive oil
substrate and 2.0 ml of deionized water was added, and
the flask was incubated for 30 min at 37C. The
reaction was stopped by adding 3.0 ml 95% (v/v)
ethanol and titrated immediately with 0.05 M sodium
hydroxide using phenolphthalein as an indicator. One
unit (U) of lipase releases one micromol of fatty acid in
one minute under the specified conditions, and the
activities were reported as Ug1).
The reaction was monitored by semiquantitative thin
layer chromatography (TLC), using Kieselgel 60 F254
TLC plates and ethyl acetate-n-heptane (4:96 v/v) as
solvent. The plates were developed for 45 min, sprayed
with a mixture of acetic acid-sulfuric acid-anisaldehyde
(100:2:1 v/v), dried for 10 min at room temperature
and heated at 105C for 5 min.
Quantitative analyses were carried out with high performance gel permeation chromatography, using an RIdetector and Ultrastyragel 500 and 100 columns
(Waters, Milford, USA) with a 0.45 mm filter in the
front of the detector. Acetone was evaporated off from
1000 ml samples in 4 h in a vacuum oven at 2.6 kPa,
after which 1000 ml HPLC-grade tetrahydrofuran was
added. The chromatograms were developed with HPLCgrade tetrahydrofuran. The components were eluted on
the basis of their molar mass in descending order. The

J BT,889-892,446 Linko

4/12/97 2:05 pm

Page 891

Production of trimethylolpropane esters of rapeseed oil fatty acids by immobilized lipase

11111
2
3
4
5
6
7
8
9
10111
1
2
3
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
5
6
7
8
9
50111
1
2
3111

rapeseed oil conversion was reported as % trimethylolpropane tri-ester or as a total conversion in % to

trimethylolpropane mono-, di- and tri-esters.
Results and discussion
Because preliminary experiments had suggested that the
conversion of rapeseed oil methyl ester to the desired
trimethylolpropane tri-esters might increase with the
use of an immobilized lipase, several carrier materials
for the C. rugosa lipase employed were investigated. The
highest total conversions of about 95% to trimethylolpropane esters were obtained only in 24 h (47C,
5.3 kPa, 13% water) with the lipase immobilized on
Duolite ES-561 (40%, w/w, biocatalyst). The highest
yield of about 70% trimethylolpropane tri-esters was
reached in 78 h (Fig. 1). When the carriers Duolite ES762, GDC 200, GCC and HPA were used relatively
high conversions were also obtained, while the other
carrier systems yielded inferior results.
With the commercial immobilized lipase, Lipozyme IM
20 (20% w/w) a conversion to trimethylolpropane triesterss was about 75% in 24 h (58C, 5.3 kPa, no added
water), while total conversions of as high as 92.5% could
be obtained. When the temperature was decreased to
47C again with no water addition, at best about 50%
conversion to trimethylolpropane tri-esters was obtained
already in 24 h, 84% in 48 h and 90% in 66 h with
40% (w/w) of the biocatalyst. The water content of the
reaction mixture had little influence to the result. Fig.
2 shows a typical time course of the transesterification
with Lipozyme IM 20 as the biocatalyst.
A stepwise addition of lipase did not markedly improve
the result. the transesterification between trimethylolpropane and rapeseed oil methyl ester, and with
Lipozyme IM 20 as the biocatalyst the product ester
profile was little affected by water content of the reaction mixture up to 15.9% water. The immobilized
lipase Novozyme 435 from C. antarctica behaved quite
differently. It was much more sensitive to the water
content than Lipozyme IM 20 from R. miehei, and typically only mono- and di-esters were formed even in 66 h
(58C, 5.3 kPa, 40%, w/w biocatalyst), with at most
traces of tri-esters. In all cases the conversion to the triester in 66 h with Lipozyme IM 20 was in excess of
80%. It should be noted here that according to a recent
report of Wehtje and Adlercreutz (1997) lipase activity
profiles as the function of water activity vary with
different lipases. This agrees well with our results. The
results also clearly showed that different immobilized
lipases behaved quite differently in the transesterification of trimethylolpropane.

Figure 1 Time course of transesterification between

trimethylolpropane and rapeseed oil methyl ester, catalyzed
by C. rugosa lipase immobilized in Duolite ES 561 (47C,
5.3 kPa, 40%, w/w biocatalyst, 15% water of substrates; r
trimethylolpropane tri-ester, u trimethylolpropane di-ester, D
trimethylolpropane mono-ester, 7 rapeseed oil methyl
ester; s unidentified compound).

Figure 2 Time course of transesterification between trimethylolpropane and rapeseed oil methyl ester, catalyzed
by immobilized R. miehei lipase Lipozyme IM 20 (58C,
5.3 kPa, 40%, w/w biocatalyst, no added water; r trimethylolpropane tri-ester, u trimethylolpropane di-ester, D trimethylolpropane mono-ester, rapeseed oil methyl ester;
sunidentified compound).

Conclusions
Total conversions to trimethylolpropane esters in the
excess of 90%, and to trimethylolpropane tri-ester of
about 75% were obtained under a reduced pressure in
24 h (58C, 5.3 kPa, no added water) with the commercial immobilized lipase Lipozyme IM 20 (20% w/w). In
66 h a conversion to trimethylolpropane tri-esters of
higher than 90% was reached. With the Candida rugosa
lipase immobilized on Duolite 561 an about 70%
conversion was obtained in 78 h at 47C, 5.3 kPa. The
optimal water content was in this case about 13%.
Relatively high conversions were also obtained when
Duolite ES-762, GDC 200, GCC and HPA 25 were
used as carriers, while with the other carriers tested inferior results were obtained. With the commercial immobilized Candida antarctica lipase Novozyme 435 only
mono- and di-esters were obtained, with traces of the
tri-ester at low water levels.
Biotechnology Techniques Vol 11 No 12 1997

891

J BT,889-892,446 Linko

4/12/97 2:05 pm

Page 892

Y.-Y. Linko et al.

11111
2
3
4
5
6
7
8
9
10111
1
2
3
4
5
6
7
8
9
20111
1
2
3
4
5
6
7
8
9
30111
1
2
3
4
5
6
7
8
9
40111
1
2
3
4
5
6
7
8
9
50111
1
2
3111

References
Bongardt, F, Bossmann, B, Westfechtel, A and Giede, W (1996).
Ger Offen. DE 4 444 137.
Carr, DD and DeGeorge, N (1989). U S Patent 4 826 633.
Coleman, MH and Macrae, A R (1977). Brit Patent 1 577 933.
Cooley, SD and Slovinsky, M (1961). U S Patent 2 991 297.
Eychenne, V, Mouloungui, Z and Gaset, A (1996) Ol, Corps Gras,
Lipides 3(1), 5763.
Leleu, G, Bdague, P and Silion, B (1977). U S Patent 4 061
581.
Linko, Y-Y, Lms, M, Huhtala, A and Linko, P (1994). J Am
Oil Chem Soc 71, 14111414.
Linko, Y-Y, Lms, M, Huhtala, A and Rantanen, O (1995a). J
Am Oil Chem Soc 72, 12931299.
Linko, Y-Y and Seppl, J (1996). CHEMTECH 26(8), 2531.
Linko, Y-Y, Wang. Z-L and Seppl, J (1995b). Enz Microb
Technol 17, 506511.
Linko, Y-Y, Wang, Z-L, Uosukainen, E, Seppl, J and Lms,
M (1996). VTT Symposium 163, 237242.

Linko, Y-Y and Wu, XY (1996). J Chem Technol Biotechnol 65,

163170.
Lms, M (1995). Finnish Patent FI 95,367.
Lms, M, Linko, P, Linko, Y-Y and Uosukainen, E (1995).
Finnish Patent FI 95,395.
Mang, T (1994). Environmentally friendly biodegradable lube
base oils technical and environmental trends in the European
market. In: Advances in Production and Applications of Lube Base
Stocks, Proceedings of International Symposium, H Singh and T S
R Prasada Rao, eds pp 6680, New Delhi, India: Tata
McGraw-Hill, 1994.
Monot, F, Benoit, Y, Ballerini, D and Vandecasteele, J-P (1990).
Appl Biochem Biotechnol 24/25, 375386.
Osada, F, Kobayashi, M, Asaoka, S and Kitazato, H (1987). Jpn
Kokai Tokkyo Koho JP 62 296 884.
Wehtje, E and Adlercreutz, P (1997). Biotechnol Lett 11, 537540.
Yokozeki, K, Yamanaka, S, Takinami, K, Hirose, Y, Tanaka, A,
Sonomoto, K and Fukui, S (1982) Eur J Appl Microbiol
Biotechnol 14, 15.

Received 20 August 1997;

Revisions requested 29 August 1997 and 25 September 1997;
Final Revisions received 10 October 1997;
Accepted 13 October 1997

892

Biotechnology Techniques Vol 11 No 12 1997