Biocatalytic enantiomeric resolution of L-menthol from an eight isomeric menthol mixture through transesterification

Article abstract of DOI:10.1016/j.molcatb.2011.10.011

The four diastereomers of menthol and their enantiomers, namely dl-menthol, dl-neomenthol, dl-neoisomenthol and dl-isomenthol, were synthesised by the hydrogenation of thymol to yield an eight isomer liquid menthol. A suitably selective lipase was sought to preferentially esterify l-menthol in organic solvent, hence simplifying separation from the diasteromeric mix through distillation. From an initial screen a commercial Pseudomonas fluorescens lipase (Amano AK) was selected, and vinyl acetate was chosen as a suitable irreversible acyl donor for transesterification. The reaction reagent ratios were optimised, and an enantiomeric excess (ee) of l-menthol of greater than 95% was reproducibly achievable at a conversion of 30% dl-menthol (0.68 M) at ≤50°C. On the basis of the composition of liquid menthol the reaction had a diastereomeric excess (de) of 82%. The enzyme was recycled 150 times in 5 ml batch reactions using liquid menthol and achieving an overall yield of 184.3 g dl-menthol/g commercial enzyme preparation. The by-product acetic acid, formed by hydrolysis of menthyl acetate, was found to cause a high degree of enzyme inactivation. The resolution reaction was scaled up 400 fold to 2 L and the enzyme recycled 38 times with an average conversion of the available l-menthol of 59% and a volumetric productivity of 1.2 g/L/h.

Full text of DOI:10.1016/j.molcatb.2011.10.011

Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
Contents lists available at SciVerse ScienceDirect
Journal of Molecular Catalysis B: Enzymatic
journal homepage: www.elsevier.com/locate/molcatb
Biocatalytic enantiomeric resolution of l-menthol from an eight isomeric
menthol mixture through transesterification
D. Bradya,∗, S. Reddya, B. Mboniswaa, L.H. Steenkampa, A.L. Rousseaua,b, C.J. Parkinsona, J. Chaplina,
R.K. Mitraa, T. Moutlanaa, S.F. Maraisa, N.S. Gardinera
a CSIR Biosciences, Private Bag X2, Modderfontein, 1645, South Africa
b Molecular Sciences Institute, School of Chemistry, University of the Witwatersrand, PO Wits 2050, South Africa
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 March 2011
Received in revised form 25 October 2011
Accepted 26 October 2011
Available online 3 November 2011
The four diastereomers of menthol and their enantiomers, namely dl-menthol, dl-neomenthol,
dl-neoisomenthol and dl-isomenthol, were synthesised by the hydrogenation of thymol to yield
an eight isomer liquid menthol. A suitably selective lipase was sought to preferentially esterify
l-menthol in organic solvent, hence simplifying separation from the diasteromeric mix through
distillation.
From an initial screen a commercial Pseudomonas fluorescens lipase (Amano AK) was selected, and
vinyl acetate was chosen as a suitable irreversible acyl donor for transesterification. The reaction reagent
ratios were optimised, and an enantiomeric excess (ee) of l-menthol of greater than 95% was reproducibly
achievable at a conversion of 30% dl-menthol (0.68 M) at ≤50 ◦C. On the basis of the composition of liquid
menthol the reaction had a diastereomeric excess (de) of 82%. The enzyme was recycled 150 times in 5 ml
batch reactions using liquid menthol and achieving an overall yield of 184.3 g dl-menthol/g commercial
enzyme preparation. The by-product acetic acid, formed by hydrolysis of menthyl acetate, was found to
cause a high degree of enzyme inactivation.
Keywords:
Transesterification
Lipases
Diasteromeric
Enantiomeric resolution
BSTR
Pseudomonas fluorescens
The resolution reaction was scaled up 400 fold to 2 L and the enzyme recycled 38 times with an average
conversion of the available l-menthol of 59% and a volumetric productivity of 1.2 g/L/h.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
non-stoichiometric ratios. Hence a method for isolation of l-
menthol from this mixture is required.
Menthol is a compound utilised in a wide range of consumer
products for its cooling effect and refreshing flavour. The demand
for l-menthol in flavour, fragrance, pharmaceutical, tobacco and
oral hygiene industries was estimated at 12,000 metric tons in
2000 [1]. The majority of the world’s supply of natural l-menthol
is derived from distillations from the leaves of the numerous sub-
species of mint (Mentha arvensis or piperita). However, menthol is
also obtained synthetically via a number of routes. One of these
is the Takasago process starting from myrcene [2], while another
is the Haarmann & Reimer/Symrise process via hydrogenation of
thymol. Unfortunately this latter reaction provides a mixture of
cyclic compounds of which l-menthol is only one. The hydro-
genation of thymol generates three stereogenic centres, and hence
yields 8 isomers, consisting of a racemic mixture of each of men-
thol, isomenthol, neomenthol and neoisomenthol (Scheme 1) in
Lipases are capable of resolving alcohol racemates through
hydrolysis or condensation of the corresponding esters. Such lipase
catalysed reactions have become particularly popular because
lipases are readily available from commercial sources, are rela-
tively inexpensive, have no co-factor requirement, and are widely
used in industry [3]. In enantiomeric kinetic resolution reactions
lipases have been demonstrated to effectively transform l-menthol
without converting the corresponding d-enantiomer. For instance,
researchers were able to isolate l-menthol from dl-menthol by
esterification with vinyl acetate or vinyl propionate in the presence
of an organic solvent [4,5], or by using tributyrin or triacetin as acyl
donors [6]. Physical isolation of l-menthyl ester from the unester-
ifed d-menthol (e.g. by distillation) followed by chemical hydrolysis
yields the desired pure l-menthol. However, the technique was not
extended to the more complex mixture of diastereomers.
Herein we describe the selective acylation of l-menthol from
the 8 isomer liquid menthol mixture by means of lipase catal-
ysed transesterification in organic solvent (Scheme 2). During
this reaction, acetaldehyde was generated from the keto-enol
∗
Corresponding author. Tel.: +27 11 605 2700; fax: +27 11 608 3020.
E-mail address: DBrady@csir.co.za (D. Brady).
1381-1177/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcatb.2011.10.011

2
D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
Lipase preparations. Pseudomonas fluorescens lipase was pur-
chased from Biocatalysis (UK) or Fluka. Horse liver esterase and
Pseudomonas cepacia lipase were obtained from Fluka. The prepa-
rations lipase AS, lipase G, lipase AYS, lipase FA15, lipase PS (P.
cepacia), and lipase AK (P. fluorescens) were all sourced from Amano
Enzyme Inc. (Japan). Wheat germ lipase, Chirazyme L4 and L6 were
purchased from Boehringer Mannheim (Germany), ESL-001-01 was
from Diversa (USA), and Carboxylesterase NP was kindly provided
by DSM (Netherlands). Other enzymes used are listed in the text or
as published previously [10].
2.1. Analytical methods
Quantitative Gas Chromatography (GC). A Stabilwax-DA wax col-
umn 30 m × 0.25 mm internal diameter (Restek, South Africa) was
mounted in a Hewlett Packard 5890A gas chromatograph. The tem-
perature profile was set as hold for 1 min at 70 ◦C, an increase of 5 ◦C
per min for 8 min to 145 ◦C holding for 1 min. Injection port and
flame ionisation detector were set at 250 ◦C. Detector gases: air at
250 kPa, H2 at 110 kPa. Auxiliary gas: N2 at 28 cm3/min. With each
set of samples a standard of l-menthyl acetate and dl-menthol was
run. Run time was 17 min. Elution times (min) vinyl acetate 1.83;
acetic acid 9.8; menthyl acetate 12.5; dl-menthol 14.2. Results are
reported as percentage conversion of available dl-menthol:
menthyl acetate
× 100 = %Conversion (mass/mass)
menthol + menthyl acetate
Scheme 1. Menthol diastereomers generated by hydrogenation of thymol.
Chiral GC method. Chrompack WCOT Fused Silica Column
25 m × 0.25 mm i.d. coating CP Chirasil-Dex CB, 0.25 ␮m film thick-
ness. Carrier gas H2 at 8 psi back pressure. Injection port and
flame ionisation detector were set at 250 ◦C. Detector gases: air
at 250 kPa, H2 at 110 kPa. Auxiliary gas: N2 at 28 cm3/min. Oven
100 ◦C (isothermal). Split vent flow: 100 cm3/min, septum purge:
3 cm3/min. Run time: 30 min. The difference in concentration of
the two menthyl acetate enantiomers as a percentage of the total
menthyl acetate provided the enantiomeric excess (%eep), of the
product menthyl acetate. This, in combination with the conversion,
was used to calculate the enantiomeric ratio of the reaction. Elu-
tion times were (in min): l-neomenthyl acetate 12.5; l-isomenthyl
acetate 13.5; d-neomenthyl acetate 13.8; d-isomenthyl acetate 14;
l-menthyl acetate 14.5; dl-neoisomenthyl acetate 16.5 and 17; d-
menthol 18; d-neomenthol 18.8; l-neomenthol 19.5; d-menthol
21.5; l-menthol 22.5; dl-neoisomenthol 22.5; l-isomenthol; d-
isomenthol. The contribution by neoisomenthol to the peak at
22.5 min was subtracted from the l-menthol peak using the data
from the non-chiral GC method.
tautomerisation of vinyl alcohol which is released from the vinyl
acetate. The equilibrium for this reaction is far towards the
acetaldehyde (Keq 3 × 10−7 at 25 ◦C) [7], making the overall reac-
tion practically irreversible and shifting the reaction equilibrium to
completion.
2. Experimental
Chemicals. Bulk l-menthol (99.7%), dl-menthol (98%) and l-
menthyl acetate (99.7%) were obtained from Haarman and Reimer
(Symrise), Germany. Vinyl acetate (99%) was obtained from
Makeean Polymers, SA; heptane (98%) from Servochem, SA. Isooc-
tane, hexane, cyclohexane, and benzene were all of analytical grade
(Servochem, SA). Other chemicals were supplied by Sigma–Aldrich
or Fluka.
Liquid menthol was generated in-house by hydrogenation
of thymol [8,9] and contained four diastereomeric pairs of
menthols, namely dl-menthol (51%), dl-isomenthol (14%), dl-
neomenthol (29%) and dl-neoisomenthol (2%), as well as 4%
menthones.
(l-menthyl acetate − d-menthyl acetate)
× 100 = %eep
(l-menthyl acetate + d-menthyl acetate)
Analytical standards included l-menthyl acetate (98% ex Aldrich
Cat. No. 44, 105-8); l-menthol (99% Fluka Cat. No. 63660); d-
menthol (>99% ex Fluka Cat. No. 63658); d-isomenthol (100%
Aldrich Cat. no. 1-7757).
Stock solutions of 40% (v/v) liquid menthol were prepared from
400 ml liquid menthol mixed with variable amounts of vinyl acetate
(depending on the experiment), which was made up to a litre with
heptane.
Scheme 2. Transesterification of l-menthol.

D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
3
2.2. Determination of enzyme activity using the 1-phenylethanol
transesterification assay (Amano standard assay).
To reduce volatility of acetaldehyde or vinyl acetate during
preparation, tubes were placed in ice water.
Enzyme, 100 mg, was weighed out in quadruplicate into glass
vials, to which 3 ml of a stock solution (20% (v/v) 1-phenylethanol
and 80% (v/v) vinyl acetate) was added. The test tubes were incu-
bated at 30 ◦C for 20 min with stirring. The reaction was stopped
by placing the test tubes in ice water. The reaction products were
then separated from the enzyme by centrifugation at 3000 rpm
for 15 min. A sample of 0.2 ml of the supernatant was added to
0.8 ml acetone in a vial and then analysed for 1-phenylethanol and
1-phenethyl acetate by quantitative GC.
(i) Menthyl acetate. Duplicate reactions containing 20 mg/ml
Amano AK, 0.7 M vinyl acetate and 0.32 M l-menthol were
performed. The dl-menthyl acetate concentration in the
6 h reaction at 25 ◦C was varied from 0 to 0.64 M. Control
experiments were prepared as above but without enzyme
addition. The effect of this constituent was also evaluated
using the Amano assay, where 1-phenylethanol (20%, v/v) was
mixed with the menthyl acetate (0–0.64 M) and vinyl acetate
(1.6–1.32 ml).
(ii) Vinyl acetate. Lipase (151 mg) was incubated with heptane and
0–3.5 M vinyl acetate (both dried over anhydrous magnesium
sulphate) at 50 ◦C for 5 h.
1-phenethyl acetate
× 100 = %Conversion
1-phenylethanol + 1-phenethyl acetate
Amano AK P. fluorescens lipase. A comparison of lipase batches
using the p-nitrophenyl palmitate (p-NPP) assay, the Amano 1-
phenylethanol assay and the menthol transesterification reaction
showed variation in enzyme activity between batches. Hence
adjustments in reaction enzyme load were made in the vari-
ous reactions to compensate for this. The p-nitrophenyl palmitate
hydrolysis assay was performed according to [11]. Lot LAKY05515
achieved 6.13% conversion (%C) by the 1-phenylethanol assay
and exhibited 27,152 U/g by the Japanese Industrial Standard
(JIS) method); Lot LAKY0950502 4.82%C (25,800 U/g by JIS); Lot
LAKV07510 10.3%C (29,000 U/g by JIS); Lot LAKX09510 5.5%C.
(iii) Acetaldehyde. Lipase (100 mg) was added to glass reaction vials.
The 1.6 M acetaldehyde stock solution consisted of 45.5 ml
acetaldehyde made up to 500 ml with heptane. Acetaldehyde
has a boiling point of 21 ◦C, and so to maintain the acetalde-
hyde in solution during the incubation period, all test tubes
were held at 15 ◦C with agitation over an extended period of 2
weeks.
(iv) Acetic acid. Lipase (20 mg) was incubated with 0–0.4 M acetic
acid in heptane at 25 ◦C.
2.7. Reaction enantioselectivity
2.3. Screening for lipase catalysed esterification of l-menthol
The effect of the presence of the other 6 isomers (dl-isomenthol,
dl-neomenthol and dl-neoisomenthol) was determined by com-
paring 40% (m/v) of the 8 isomer mix as starting material to
reactions using 10% synthetic dl-menthol. After addition of vinyl
acetate (2.65 ml) the reaction was topped up with heptane to 10 ml.
Enzyme loads of 0.3 and 0.6 were compared at 40 ◦C with stirring
over four reaction cycles in Multi-ReactorTM (Robo Synthon, CA,
USA). These consisted of jacketed reactor vessels that housed up
to 16 × 50 ml test tubes, and was fitted with a temperature control
unit and a magnetic stirrer. The test tubes were closed with a Teflon
cap.
Isomenthol enrichment was achieved using a dl-isomenthol
and dl-menthol mixture distilled from liquid menthol, and then
adding dl-menthol to achieve various ratios of dl-isomenthol and
dl-menthol.
The influence of the ratio of d-menthol to l-menthol was inves-
tigated by blending dl-menthol and l-menthol. The reactions were
performed with 26% (m/m) total menthol, 337.5 mg Amano AK, and
1 equiv. of vinyl acetate with respect to total menthols present.
In 2 ml vials 1–10 mg lipase was incubated at 30 ◦C in 2 ml
organic solvent (toluene, cyclohexane, hexane, pentane, decane,
2-propanol or heptane) containing 0.075 M l-menthol and 0.075 M
of each acyl donor (vinyl acetate, butyric acid, octanoic acid or lau-
ric acid). The reaction was stopped by centrifuging the enzyme
out of suspension (3000 rpm for 15 min) and a sample from the
supernatant spotted onto TLC plates for analysis. TLC analysis was
performed on silica-60 plates using 90:10 hexane:ethyl acetate as
the mobile phase. The plates were dipped in 0.5% KMnO4 in acetone,
and heat dried for spot colour development.
2.4. Comparison of Amano AK and PS lipases
Reactions (20 ml) were performed at 40 ◦C in the carousel reac-
tion system (Radleys Starfish, Radleys, UK). Samples were taken
after 24 h and analysed for menthol (% mass/mass) and for the ratio
of d-menthol to l-menthol.
2.5. Reaction kinetic studies
2.8. Reaction optimisation
To determine the influence of vinyl acetate concentration on
reaction performance, duplicate reactions containing increasing
concentrations of vinyl acetate (0–2.6 M) were conducted. An l-
menthol concentration of 3.2 M (50%, m/v) was used with heptane
as solvent. Amano AK enzyme, 151 mg, was added to each 5 ml reac-
tion. All reactions were performed for 5 h at 50 ◦C by incubation in
stirred silicon oil baths on magnetic stirrer hot plates (750 rpm).
A second set of reactions were performed under the same condi-
tions with increasing concentrations of l-menthol 0–3.2 M (0–50%,
m/v) and a fixed vinyl acetate concentration of 3.2 M.
A statistically designed set of experiments was drawn up using
the Design Ease software package (Stat-Ease Inc., MN, USA) in order
to study the effect of variables on the conversion and enantiose-
lectivity of the lipase catalysed esterification of dl-menthol. A 24
(half factorial) design was drawn up with ranges for four vari-
ables: temperature, dl-menthol and vinyl acetate concentration,
and enzyme loading (mg of enzyme per g of dl-menthol). Each of
the reactions of the statistical design was carried out on a 250–300 g
scale in the Maxi-ReactorTM (Robo Synthon, CA, USA) that consisted
of 6 × 450 ml jacketed reactor vessels, with individually controlled
temperature and agitation units. The agitation was provided by
bi-convex-shaped magnetic stirrer bars.
2.6. Enzyme inactivation by reaction components
The enzyme was exposed to reactants or products, washed three
times with heptane to remove the compound (with recovery by
centrifugation at 3000 rpm for 15 min), and the residual activity of
the enzyme determined by the Amano assay.
2.9. Optimum molar ratio of vinyl acetate to l-menthol
Amano AK lipase (151 mg) was added to glass reaction vials.
For each of the ratios evaluated, the dl-menthol concentration in

4
D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
Table 1
Screening of selected enzymes for esterification of menthol with mixed acyl donors.
Enzyme
Solvent
Acyl donor
Toluene
Cyclohexane
Hexane
Pentane
Decane
2-Propanol
Heptane
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
−
ND
ND
ND
ND
+
ND
ND
ND
ND
−
+
+
−
ND
ND
ND
ND
−
+
Hog pancreas lipase
Fluka
Cat No. 62300
++
++
++
−
−
−
−
−
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
ND
ND
ND
ND
ND
ND
ND
ND
−
−
−
−
+
−
ND
ND
ND
ND
Hog liver Esterase
Boehringer
−
ND
ND
−
++
−
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
−
−
++
+
−
−
−
+
ND
ND
ND
ND
Porcine pancreas
Chirazyme L6
Boehringer
+
++
−
++
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
−
−
+
ND
ND
ND
ND
+
ND
ND
ND
ND
−
+
+
+
ND
ND
ND
ND
−
−
−
+
Aspegillus oryzae lipase
Fluka
Cat No. 62285
−
+
++
−
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
+
−
+
+
++
−
−
−
−
−
−
++
+
−
−
+
Rhizomucor miehei lipase
Fluka
Cat No. 62291
−
−
+
−
ND
ND
ND
++
+
−
ND
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
−
++
−
−
−
+
−
+
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Rhizopus japonicus
Nagase enzyme lipase A-10FG
−
++
−
++
++
+
++
++
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
+
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Candida antarctica lipase
Fluka
Cat No. 62299
++
−
−
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
++
−
+
−
++
−
ND
ND
ND
ND
ND
ND
ND
ND
+
−
−
+
−
ND
ND
ND
ND
Candida antarctica fraction A
Chirazyme L5
Boehringer
−
−
+
++
−
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
−
+
+
+
−
+
+
−
+
+
+
−
−
−
+
ND
ND
ND
ND
Candida cylindracea lipase
Fluka
Cat No. 62316
−
−
++
−
−
−
−
++
−
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
++
++
−
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Chromobacterium viscosum
lipoprotein lipase
Fluka
+
Cat No. 62333
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
ND
++
+
+
+
++
−
+
+
+
−
+
+
ND
ND
ND
ND
Burkholderia Chirazyme L1
Boehringer
−
++
−
−
−
−
−
+
+
−
++
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
++
−
ND
ND
ND
ND
−
−
−
−
+
+
−
ND
−
ND
ND
ND
ND
Pseudomonas cepacia
lipoprotein lipase
Fluka
−
++
−
−
++
−
+
−
−
Cat No. 62309
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
−
−
−
+
−
++
+
−
ND
ND
ND
ND
+
ND
ND
ND
ND
−
+
Pseudomonas sp. Chirazyme L4
Boehringer
−
+
−
++
−
+
+
++
−
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
−
−
+
++
−
ND
ND
ND
ND
ND
ND
ND
ND
−
+
+
−
−
−
−
+
+
+
+
Pseudomonas. fluorescens lipase
Fluka
Cat No. 62312
−
++
−
++
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
−
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
−
+
Recombinant ex Bacillus thai
carboxylesterase NP
DSM
−
ND
−
−
−
Metagenome
ESL-001-01
Recomb. Biocat.
Vinyl acetate
Butyric acid
Octanoic acid
Lauric acid
+
−
+
+
+
+
+
+
++
−
−
+
+
−
++
−
ND
++
ND
−
−
−
ND
++
ND
ND
++
ND
“−”, “+” and “++” represent no activity, some activity, and strong activity, respectively.
ND = Not Determined.

D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
5
Table 2
A comparison between the performance of Amano lipases AK and PS on the esterification of dl-menthol at 24 h.
Reaction cycle
Enzyme
Substrate
concentration (%)
Conversion (%)
ee (%)
E
Menthol mol
balance (%)
None
PS
PS
PS
PS
40
10
25
40
10
25
40
10
25
40
10
25
40
0.1
0.8
1.3
1.5
1.0
1.5
1.5
5.5
13.1
15.2
10.3
20.5
21.3
0
1.0
6.8
8.8
8.4
7.2
8.9
9.0
30.0
51.2
52.6
43.4
62.4
64.8
100
104
108
111
104
104
106
101
106
109
102
104
101
1
1
1
2
2
2
1
1
1
2
2
2
74.2
79.4
78.4
75.4
79.6
79.8
93.2
95.6
95.6
95
PS
PS
AK
AK
AK
AK
AK
AK
96
96.1
heptane was maintained at 2.28 M (made up to 20 ml) and the vinyl
acetate to l-menthol molar ratio was varied from 0.25:1 to 3:1. The
test tubes were incubated at 50 ◦C. Each resolution was performed
in quadruplicate.
enzyme with octanoic acid in decane gave, at best, 26.2% conversion
and an ee of only 28.6% (an E of 1.9). Hence investigation into this
enzyme was discontinued. Animal extract derived enzymes were
not investigated further due to considerations relating to use for
food products.
2.10. Small scale enzyme recycle
Three of the positive reactions were catalysed by enzymes
from Pseudomonas species, which are known to be selective for l-
menthol [4,5], while two of the preparations were from Candida
species, which are also known to be selective for this substrate
[12]. Comparison of Pseudomonas sp. lipases from different sup-
pliers during the esterification of menthol by vinyl acetate was
performed. P. cepacia lipase from Fluka and Amano showed differ-
ences in the esterification of menthol (12 and 17%, respectively). P.
fluorescens lipase from Biocatalysis did not esterify menthol, while
P. fluorescens lipase preparation from Fluka did (19%). The Pseu-
domonas lipase L4 from Boehringer was less effective (12%). Such
differences could be due to the strains employed by various manu-
facturers. Due to their availability in industrial quantities, crude
preparations of P. fluorescens lipase (Amano AK) and P. cepacia
(Amano PS) were evaluated further.
Glass test tubes (10 ml) with solvent resistant Teflon caps and
containing magnetic stirrer bars were used for small scale batch
recycles. To the tubes were added 5 ml of a 20% menthol stock
solution and 100 mg of the enzyme. The initial vinyl acetate to
l-menthol ratio for the transesterification reaction was 2:1. Incu-
bation was at 50 ◦C for 24 h.
2.11. Batch stirred tank reactions
Thermally insulated glass vessels (2.2 L) with water heating
jackets and glass baffles (Glasstek, South Africa) were connected to
circulating water baths set at 50 0.5 ◦C. The reactors were charged
with 2 L of reaction mixture, which consisted of vinyl acetate 5.6%
(v/v) (2:1 molar ratio to l-menthol), liquid menthol 20% (v/v),
heptane (to volume) 74.4% (v/v), plus 25 g/L enzyme (Amano AK,
Lot LAX09510). Agitation at 450 rpm for 23 h was achieved with
magnetic stirrer bars or Heidolph R2R 2021 overhead stirrers fit-
ted with 2 radial flow impellers, thereby keeping the enzyme in
suspension.
Filtration was more practical than centrifugation at large scale,
and hence the stirrers were stopped to let the enzyme settle from
suspension, and a sintered filter (porosity 3: 16–40 ␮m pore size,
6 cm diameter) was suspended in the organic phase. This was con-
nected to a vacuum pump (at 60–80 kPa) and used to remove
the liquid. The enzyme was washed with 100 ml heptane and
then re-charged with reaction mixture for the next reaction cycle.
The heptane from the wash step was combined with the product
mixture.
3.2. Comparison of Amano AK and PS lipases
The P. cepacia (PS) lipase from Amano has previously been
demonstrated to have a high enantioselectivity towards l-menthol
in this and similar reactions [5,13]. However, in 20 ml reactions
involving dl-menthol and vinyl acetate, the Amano AK enzyme
resulted in a higher enantiomeric excess of l-menthol (Table 2), and
was therefore selected for all further studies. Heptane was selected
as the solvent of choice due to low neurotoxic properties [14] and
a suitable partition coefficient [15].
The optimum Amano AK lipase concentration required to
produce almost 50% conversion of 12% (v/v) dl-menthol was estab-
lished in 5 ml reactions. The enzyme concentration of 1.2 g per %
liquid menthol in a litre reaction (Fig. 1) provided near complete
conversion and was chosen for future experiments.
3. Results and discussion
3.3. Reaction kinetic studies
3.1. Screening for lipase catalysed esterification of l-menthol
For vinyl acetate (the co-substrate), the enzyme KM was deter-
mined to be 0.33 M (2.8%, m/v). Therefore to avoid acyl donor based
kinetic limitations, enough vinyl acetate should be included in the
initial reaction mixture to leave a residual of 0.7 M (5.6%) upon
complete conversion of the l-menthol.
The KM value was 1.13 M for l-menthol (17.6%, m/v) (Fig. 2).
This indicates that the enzyme activity would be limited by the l-
menthol concentration throughout the transesterification reaction
of 40% (m/v) liquid menthol. The maximum specific activity of the
Enzymes were evaluated for the formation of menthol esters in
the presence of a mixture of acyl donors (Table 1). The promising
thermostable enzyme ESL-001-01 was evaluated at the 1 ml scale
(with either vinyl acetate or octanoic acid as an acyl donor) in a
range of solvents and at three temperatures (30, 50 or 70 ◦C), but
could not provide an E above 7 (data not shown). When vinyl acetate
in combination with toluene was used at 30 ◦C, an ee of 74.6% was
achieved but at a conversion of only 1.4% and an E of 6.9. The same

6
D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
Fig. 3. Residual activity of Amano AK (as measured by the Amano assay) after expo-
sure to increasing acetic acid concentrations in heptane at 25 ◦C.
Fig. 1. Transesterifcation of 12% (v/v) liquid menthol with varying enzyme concen-
trations.
(iv) Acetic acid. Acetic acid formation was previously detected in the
transesterification reactions. This experiment indicated that
addition of acetic acid had a significant inactivating effect on
the enzyme (Fig. 3). Over 20% of enzyme activity was lost
after incubation for 20 h in 0.02 M acetic acid. No residual
enzyme activity was displayed after exposure to 0.2 M acetic
acid under these conditions. As this dramatic loss of activity
was not observed to occur with incubation with vinyl acetate,
it can be assumed that acetic acid formation occurs through
the water-mediated hydrolysis of menthyl acetate. Available
sources of water for the reaction are the water associated with
the biocatalyst (up to 10% (m/m) in spray- and freeze-dried
enzyme preparation) and a low level of water in the reaction
components and bulk solvent. Due to the severity of the dena-
turing effect of acetic acid on the activity of the enzyme, the
production of acetic acid, and therefore water addition, must
be limited under process conditions. At a reaction tempera-
ture of 50 ◦C the rate of denaturation in the presence of acetic
acid could be anticipated to completely denature the enzyme.
However, the biocatalyst exhibits significantly greater stabil-
ity under process conditions than would be expected from this
experiment, indicating that other factors are involved.
Fig. 2. Specific activity of Amano AK with vinyl acetate (ꢀ) and l-menthol (ꢁ) con-
centration determined in excess of the co-substrate.
biocatalyst was 0.25 g menthyl acetate/g enzyme preparation/h on
l-menthol.
A comparison of reaction rates on l-menthol, dl-menthol and
liquid menthol up to l-menthol levels of 1.3 M was also con-
ducted. The reaction rates with l- and dl-menthol were similar
at comparative l-menthol concentrations, indicating no significant
interference by the d-enantiomer on the initial rate of reaction.
However, the study with liquid menthol indicated that the appar-
ent specific activity on liquid menthol was at least 40% lower
than was observed with l-menthol (data not shown). This sug-
gests that while the other menthol isomers were poor substrates
for this enzyme, some competition for binding at the active site may
occur.
3.5. Reaction enantioselectivity
To evaluate whether the other menthol isomers influenced
the enantioselectivity of the reaction, two experiments were
performed. Firstly, the isomenthol component was artificially
enhanced, but this did not influence reaction enantiomeric excess
(Table 3). Having eliminated the influence of isomenthol, it was
possible to determine the effect of neomenthol by comparison of
dl-menthol with liquid menthol (which contained a significant pro-
portion of neomenthol (29%) as well as isomenthol and traces of
isoneomenthol). As can be seen (Table 4), the presence of the other
6 isomers had little effect on the enantioselectivity.
Increasing the amount of d-menthol present reduced the ee of
the l-menthyl acetate formed (Table 5), falling below 95% ee when
the starting l-menthol concentration was only 25% of the total dl-
menthol. In a reaction this would occur at 30% conversion of the
dl-menthol.
3.4. Enzyme inactivation by reaction components
The enzyme was incubated with reaction matrix components,
washed with heptane, and the residual enzyme activity measured
by the Amano assay.
(i) Menthyl acetate. There appeared to be a 72% increase in activ-
ity on addition of 0.64 M dl-menthyl acetate to the reaction.
This effect was confirmed using the Amano assay with a 30%
increase in activity at the same concentration.
3.6. Reaction optimisation
(ii) Vinyl acetate up to concentrations of 3.5 M did not appear to be
toxic to the enzyme.
Analysis of the results of the statistically designed set of exper-
iments (Table 6) indicates that increased temperature had the
largest negative effect on the enantiomeric ratio (Fig. 4), resulting
from the decrease in the enantiomeric ratio at higher conversion.
This effect has been previously observed [5,16]. By comparison,
vinyl acetate concentration (or molar ratio of vinyl acetate to
(iii) Acetaldehyde. A 20% loss in residual enzyme activity was
observed after 48 h incubation in 0.3 M acetaldehyde at 15 ◦C.
Although the denaturing effect would probably increase with
temperature, so would the removal of acetaldehyde due to
volatility (in a ventilated reactor).

D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
7
Table 3
The effect of isomenthol concentration on the enantiomeric excess.
dl-Menthol (%, m/m)
Isomenthol (%, m/m)
Isomenthol (% liquid menthol)
dl-Menthol conversion (%)
ee (%)
E
40.0
49.8
49.5
46.9
47.3
22.4
0.0
2.8
4.6
9.1
10.3
19.6
0.0
5.2
8.6
16.2
17.8
47.0
20.3
17.3
17.5
17.4
20.0
17.7
96.7
96.9
96.8
97.0
97.0
97.2
75.8
77.5
75.2
80.1
83.2
86.4
Table 4
Enantioselective l-menthylacetate synthesis.
Substrate
Enzyme mass (mg)
Cycle 1
%C
Cycle 2
%C
Cycle 3
%C
Cycle 4
Average %C
Average %ee
Average E
%ee
%ee
%ee
%C
%ee
10% dl-menthol
10% dl-menthol
40% liquid menthol
40% liquid menthol
302
604
302
604
21.6
22.4
17.8
23.8
96.7
97.1
97.2
97.1
41.2
48.4
31.7
46.2
95.7
94.6
96.3
94.8
38.6
50.8
32.8
47.4
95.6
93.4
96.1
94.0
39.1
49.0
30.9
47.4
95.4
93.6
96.1
94.0
35.1
42.7
28.3
41.2
96
96
95
95
83
81
102
90
%C: conversion of l-menthol to l-menthyl acetate.
Table 5
The effect of menthol enantiomeric ratio on the conversion and enantioselectivity of the Amano AK lipase catalysed acetylation of menthol.
Initial l-menthol as
% of total
% Menthyl acetate formed (40 ◦C)
dl-menthol
8 h
l-Menthyl acetate
15.7
17.5
18.5
19.7
23.7
0
24 h
d-Menthyl acetate
%ee
100
99
98
97
91
0
l-Menthyl acetate
d-Menthyl acetate
%ee
100
0
0.09
0.2
0.3
1.07
0.83
48.15
54.5
57.4
59.75
68.4
0
0
100
99
97
96
88
0
75
60
50
25
0
0.33
0.75
1.165
4.35
2.85
menthol) only had a small positive effect on reaction enantiose-
lectivity. The data also showed that the enzyme reaction could be
performed at 50 ◦C and provide a 35% conversion of dl-menthol,
while maintaining an ee of 95%, and an E of 64.
the l-menthol, and providing a kinetic advantage towards the
end of the reaction. Although the kinetic data showed that
the enzyme had a higher affinity for vinyl acetate than for
menthol, excess vinyl acetate is required to optimise the reac-
tion rate and to compensate for volatilisation into the reactor
headspace.
3.7. Optimum molar ratio of vinyl acetate to l-menthol
Similar to the reaction optimisation study, Fig. 5 shows that
ratios of 1:1 to 3:1 vinyl acetate to menthol demonstrated sim-
ilar initial results during the transesterification of dl-menthol
to menthyl acetate, but the final product increased about 25%
at ratios of 2.5:1 and 3:1, giving about 80% conversion of
3.8. Small scale enzyme recycle
To determine the recyclability of the enzyme, the Amano AK
lipase was reused 150 times (Fig. 6) achieving an average concen-
tration of 171 mM menthyl acetate and an overall yield of 184.3 g
Table 6
Statistically designed experiments for reaction optimisation.
Run
dl-Menthol
concentration (%, m/m)
Temperature (◦C)
Enzyme (mg/g)
Vinyl acetate:menthol
ratio
dl-Menthol
conversion (%)
eep%
E
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
60
20
60
60
20
20
20
60
40
40
20
20
20
60
20
60
35
35
50
35
50
35
50
50
42.5
42.5
35
35
35
50
50
35
100
50
50
1:1
1:1
1:1
2:1
1:1
2:1
2:1
2:1
1.5:1
1.5:1
1:1
1:1
1:1
1:1
2:1
2:1
25.9
12.9
32.9
16.6
34.3
18.4
24.4
42.8
30.2
29.2
13.8
14
96.3
93.0
94.3
96.1
94.7
97.9
96.1
93.8
96.6
96.7
97.8
97.8
97.8
93.8
95.6
97.4
73.7
31.6
53.9
60.6
60.1
117.0
68.1
65.8
87.3
88.2
104.8
105.1
114.2
65.8
50
100
100
50
100
75
75
50
50
100
100
100
100
20
42.8
31.9
26
69.2
106.1

8
D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
Table 7
Bench scale batch stirred tank enzyme recycle experiments.
Reaction recycle
Conversion (%, m/m)
Std. dev.
BSTR 1
BSTR 2
BSTR 3
BSTR 4
Mean
1
2
3
4
5
6
7
8
25.67
30.64
34.45
38.30
24.49
33.45
34.37
34.37
ND
33.57
37.52
38.04
38.40
22.54
30.18
30.18
34.4
32.67
33.76
33.59
30.75
30.68
31.61
ND
29.23
29.53
ND
27.63
27.78
29.55
29.96
27.52
25.01
26.52
23.59
23.44
21.58
19.99
26.02
31.44
35.61
35.84
28.71
27.49
29.14
30.27
31.12
29.46
34.97
39.05
34.88
22.67
22.23
27.73
32.11
30.2
24.80
31.96
33.37
33.56
28.48
33.67
27.58
29.69
29.75
ND
28.53
38.74
39.25
23.68
24.04
28.88
31.91
29.8
32.42
30.25
30.27
29.2
29.65
29.28
28.41
28.62
26.01
25.97
29.5
24.53
31.94
33.50
33.84
28.40
33.39
28.70
29.63
30.47
29.49
35.33
36.78
38.14
23.90
26.00
30.77
30.13
30.74
32.47
20.65
30.78
29.15
31.24
32.09
29.04
29.19
29.43
25.73
29.81
29.92
28.8
25.25
31.50
34.23
35.39
27.52
32.00
29.95
30.99
30.44
30.84
34.09
38.15
37.67
23.20
25.61
29.39
32.14
30.85
32.46
29.54
30.06
29.52
30.55
30.69
28.57
28.88
27.94
25.99
29.31
29.31
28.92
27.96
25.92
27.30
24.10
23.24
23.33
21.87
0.71
0.62
1.03
2.19
2.03
3.01
3.02
2.27
0.69
2.36
3.87
1.01
1.92
0.69
3.41
1.36
1.75
1.27
1.05
6.13
1.11
0.78
1.02
1.99
0.73
0.60
1.75
1.24
1.05
0.53
0.74
0.36
1.85
0.59
0.95
3.01
3.43
3.42
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Fig. 4. Interaction between temperature, dl-menthol concentration, and enan-
tiomeric excess. Enzyme loading at 100 mg/g.
31.2
33.65
28.42
29.04
29.7
ND
27.61
28.18
28.37
24.62
30.14
28.75
28.71
28.19
24.7
27.28
23.02
20.97
21.91
20.4
29.01
28.2
27.8
28.67
27.45
24.84
21.12
21.38
20.08
28.31
25.31
27.94
24.96
27.43
28.46
26.99
Fig. 5. The effect of varying the molar ratio of vinyl acetate to l-menthol during
transesterification in the presence of 2.6 M (40%, m/v) dl-menthol.
Average
30.14
28.75
28.81
29.56
29.33
1.72
3.9. Bench scale batch stirred tank reactions
l-menthol/g enzyme preparation. This sustained activity was pos-
sible as lipases tend to be more stable in organic solvents than
in aqueous media [17]. The conversion for the first cycle was
lower than the initial recycle, a phenomenon that was frequently
observed, and possibly due to menthyl acetate hydrolysis caused
by unbound water associated with the enzyme [16]. This resulted
in the formation of an average of 54 mM acetic acid in the reaction
mixture.
Having determined that the ee was >95% at 30% conversion of
dl-menthol in the liquid menthol (i.e. 15.3% of the total menthols)
during reaction optimisation (above) we then ran a set of four reac-
tors in batch mode with enzyme recycle, wherein the target was
limited to 30% conversion of dl-menthol. To limit the conversion
to an average of 30% the enzyme load was reduced to 25 g/L, as cal-
culated from the data (Fig. 1). Although there is some variability
in the conversion (Table 7), this correlated with variations in the
batches of the liquid menthol and agitation. The average l-menthyl
acetate volumetric productivity was 1.2 g/L/h.
Enzyme losses occurred due to formation of airborne dust by
the enzyme powder during manipulations of the enzyme between
reaction cycles, and adhesion on the sintered glass filters during
filtration. These losses were in the order of 25 mg (0.1%) per recy-
cle. In addition a proportion of the enzyme activity was lost due
to thermal and physical denaturation. In spite of these losses, the
activity remained relatively constant over extended periods, with
an average conversion of l-menthol of 59%.
This reaction provided an enantiomeric ratio (E) of >50. Impor-
tantly when 95% of the l-menthol in the liquid menthol was
esterified, the bulk of the other menthol isomers remained unester-
fied. Only 4% of the d-menthol, 2.5% of the dl-neomenthol, and
3.5% of the dl-isomenthol were esterified. A much higher portion
of the dl-neoisomenthol (18.5%) was esterified, although this was
Fig. 6. Batch recycles of Amano AK lipase enzyme at 50 ◦C. Low conversion for
recycle 45 was due to malfunction of temperature control.

D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
9
Symrise Process
Biocatalyꢀc
Process
m-cresol
thymol
m-cresol
Alkylaꢀon
Alkylaꢀon
thymol
Hydrogenaꢀon/
Racemisaꢀon
Hydrogenaꢀon/
Racemisaꢀon
menthol isomers
Fracꢀonal disꢀllaꢀon
menthol isomers
Selecꢀve resoluꢀon by
DL-menthol
enzymaꢀc esterificaꢀon
Non-selecꢀve
esterificaꢀon
DL-menthyl
benzoate
Simplified
disꢀllaꢀon
Selecꢀve
7 menthol
isomers
crystallisaꢀon
L-menthyl
benzoate
D-menthyl
benzoate
Chemical hydrolysis
Crystallisaꢀon
Chemical hydrolysis
Crystallisaꢀon
L-menthol
L-menthol
Fig. 7. Comparison of the Symrise process and the process described here.
insignificant as dl-neoisomenthol typically comprised only 2–3%
of the liquid menthol. This yielded a diastereomeric excess (de) of
82% of l-menthol per total menthol.
seems to be reduced during operational studies, suggesting that
the substrate may stabilise or protect the enzyme. Menthyl acetate
hydrolysis can also reduce overall reaction yields on vinyl acetate
and the reaction rate, and hence the water available in the reaction
mixture must be kept to a minimum.
The reaction was successfully scaled up by a factor of 400 (from
5 ml to 2 L) in batch stirred tank reactors at 50 ◦C and the enzyme
was recycled a total of 38 times at this scale. The free enzyme can
convert 30% of the dl-menthol with an enantiomeric ratio of 65 and
a diastereomeric excess (de) of 82%.
After the reaction, the mixture can be distilled to remove the l-
menthyl acetate which can be hydrolysed in a 70:30 ethanol:water
mixture with 1.1 equiv. of sodium hydroxide and the resul-
tant l-menthol isolated by crystallisation. The toxic acetaldehyde
generated is vacuum distilled and combusted. The other 7 men-
thol isomers (and residual l-menthol) can be recycled back to
liquid menthol through a racemisation reaction, ensuring com-
plete use of the material. The menthol isomers were isomerised
and racemised, over the same nickel catalyst used for thymol
hydrogenation, in a slurry reactor at 200 ◦C and 0.6 MPa H2
[9,18].
This process contains fewer and simpler unit operations than
the original Symrise process on which it is based (Fig. 7). Another
interesting biocatalytic variation of the Symrise process involves
highly selective Candida rugosa lipase based hydrolysis of the l-
enantiomer of dl-menthyl benozoate [19].
Acknowledgements
We would like to acknowledge the technical inputs of a
large number of people who contributed to this research, includ-
ing D. Dikoko, S. Isaacs, V. Maebela, L. Boer, B. Cowan, P.
Kunene, V. Maebela, S. Moema, V. Ndovela, M. Portwig, S.J. Heg-
gie, M. Odayar, and G. Baloyi. We would also like to show
our appreciation to AECI Ltd. (South Africa) for supporting this
research.
4. Conclusion
References
In this research we demonstrated the selective esterification
of l-menthol from an 8 isomer mixture. Amano AK in heptane
catalysed the transesterification of l-menthol with vinyl acetate
at a significantly higher rate than d-menthol resulting in menthyl
acetate enriched in the l-enantiomer. The presence of isomenthol
or neomenthol had no significant effect on either the rate of men-
thol conversion or the final ee of the menthyl acetate. Although
the presence of the other isomers reduced conversion, this could
be compensated by increasing the enzyme load. The reaction by-
products acetaldehyde and, particularly, acetic acid were shown
to inactivate the enzyme. Interestingly, the strength of this effect
[1] M. McCoy, Chem. Eng. News 88 (2010) 15–16.
[2] H. Yoji, I. Takeshi, O. Yoshiki, Euro. Pat. 1,225,163 (2002).
[3] O. Kirk, T.V. Borchert, C.C. Fuglsang, Curr. Opin. Biotechnol. 13 (2002)
345–351.
[4] W.-H. Wu, C.C. Akoh, R.S. Phillips, J. Food Lipids 3 (1996) 189–198.
[5] W.-H. Wu, C.C. Akoh, R.S. Phillips, JAOCS 74 (1997) 435–439.
[6] C.J. Gray, J.S. Narang, S.A. Barker, Enzyme Microb. Technol. 12 (1990)
800–807.
[7] A.K. Cederstav, B.M. Novak, JACS 116 (1994) 4073–4074.
[8] J. Dudas, J. Hanika, J. Lepuru, M. Barkhuysen, Chem. Biochem. Eng. Q. 19 (2005)
255–262.
[9] J. Dudas, J. Hanika, Chem. Eng. Res. Des. 87 (2009) 83–90.
[10] L. Steenkamp, D. Brady, Enzyme Microb. Technol. 32 (2003) 472–477.

10
D. Brady et al. / Journal of Molecular Catalysis B: Enzymatic 75 (2012) 1–10
[11] T. Vorderwülbecke, K. Kieslich, H. Erdmann, Enzyme Microb. Technol. 14 (1992)
631–639.
[12] S.S. Othman, M. Basri, M.Z. Hussein, M.B.A. Rahman, R.N.Z.A. Rahman, A.B.
Salleh, H. Jasmani, Food Chem. 106 (2008) 437–443.
[13] E. Cernia, C. Palocci, S. Soro, Chem. Phys. Lipids 93 (1998) 157–168.
[14] Impurities: Guideline for Residual Solvents CPMP/ICH/283/95 ICH Topic Q3C
(R4) February 2009. European Medicines Agency.
[15] C. Laane, S. Boeren, K. Vos, C. Veeger, Biotechnol. Bioeng. 30 (1987) 81–87.
[16] S. Bai, G. Zheng, W. Liu, Y. Sun, Food Chem. 96 (2006) 1–7.
[17] A. Zaks, A.M. Klibanov, J. Biol. Chem. 263 (1988) 8017–8021.
[18] J.A. Chaplin, N.S. Gardiner, R.K. Mitra, C.J. Parkinson, M. Portwig, M.D. Dickson,
D. Brady, S.F. Marais, S. Reddy, US Pat. 7,026,144 (2002).
[19] U. Bornscheuer, I.-L. Gatfield, J.-M. Hilmer, R. Schmidt, S. Vorlova, Euro. Pat.
1,222,223 (2002).