Lipase-catalysed kinetic resolution of 1-O-alkylglycerols by sequential transesterification

Article abstract of DOI:10.1016/j.tetasy.2004.07.046

The natural S-configured chimyl, batyl and selachyl alcohols of the 1-O-alkylglycerol type were prepared by enantioselective lipase-catalysed transesterification. Their racemates were synthesised in two steps by reacting racemic solketal with the bromides of the corresponding fatty alcohols and a subsequent conversion of the intermediates into the 1-O-alkylglycerols by deprotection under acidic aqueous conditions. The Pseudomonas fluorescens lipase was employed to kinetically resolve the racemic 1-O-alkylglycerols by a sequential diacetylation process to afford them virtually enantiomerically pure. Dramatic enantioselectivity increase was observed for the saturated chimyl (E = 17-32) and batyl (E = 14-38) alcohols at decreased temperature, whereas for the monounsaturated selachyl (E = 12-13) alcohol no such temperature effects were observed.

Full text of DOI:10.1016/j.tetasy.2004.07.046

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 2893–2899
Lipase-catalysed kinetic resolution of 1-O-alkylglycerols
by sequential transesterification
Arnar Halldorsson, Pall Thordarson, Bjorn Kristinsson, Carlos D. Magnusson
and Gudmundur G. Haraldsson*
Science Institute, University of Iceland, Dunhaga 3, IS-107 Reykjvik, Iceland
Received 28 May 2004; accepted 9 July 2004
Abstract—The natural S-configured chimyl, batyl and selachyl alcohols of the 1-O-alkylglycerol type were prepared by enantio-
selective lipase-catalysed transesterification. Their racemates were synthesised in two steps by reacting racemic solketal with the
bromides of the corresponding fatty alcohols and a subsequent conversion of the intermediates into the 1-O-alkylglycerols by
deprotection under acidic aqueous conditions. The Pseudomonas fluorescens lipase was employed to kinetically resolve the racemic
1-O-alkylglycerols by a sequential diacetylation process to afford them virtually enantiomerically pure. Dramatic enantioselectivity
increase was observed for the saturated chimyl (E = 17–32) and batyl (E = 14–38) alcohols at decreased temperature, whereas for the
monounsaturated selachyl (E = 12–13) alcohol no such temperature effects were observed.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
O
O
H
H
OH
OH
OH
OH
Nonpolar ether lipids of the 1-O-alkyl-2,3-diacyl-sn-
glycerol type are major constituents in the liver oils of
shark, dogfish and various other species of elasmo-
branch fish.1–3 Shark liver oil has been used for a long
time as a therapeutic and preventive agent, with 1-O-
alkyl-sn-glycerols as the active constituents.
1b
1a
O
H
OH
OH
The 1-O-alkyl-sn-glycerols, sharing certain similarities
with the well-known platelet activating factors, have
been claimed to display various beneficial effects on
human health.4,5 Their early clinical use was to prevent
radiation sickness from cancer X-ray therapy.6 Results
from scientific research in recent years show that they
can be stimulating for the allergic system as well as for
the immune control, asthma, psoriasis, arthritis or to
speed up the removal of heavy metals from the body.7
1c
Figure 1. The chemical structure of the three main natural 1-O-alkyl-
sn-glycerols present in shark liver oil: chimyl 1a, batyl 1b and selachyl
1c alcohols.
liver oils of chimaeras, sharks and rays of the Chimaeroi-
dei, Batoidei and Selachoidei families.
Lipases are among the most widely applied and versatile
biocatalysts in organic synthesis. The most widespread
application of lipases has undoubtedly been in asymmet-
ric transformations, which has been reviewed compre-
hensively.9–12 Amongst the lipases the pig pancreatic
lipase has been the most widely used, but in the course
of last 20 years microbial lipases have gained increasing
importance, especially the commercially available bacte-
ria lipases from Pseudomonas. Lipases exhibit high toler-
ance for variations in substrate structure and their active
As the sn-terminology implies8 their natural absolute
configuration is S. The three most abundant fatty alcoh-
ols present in the 1-O-alkyl moiety are, C16:0, C18:0 and
C18:1, the last one being the most abundant. They corre-
spond to chimyl 1a, batyl 1b and selachyl 1c alcohols
(Fig. 1), respectively, named after their sources in the
*
Corresponding author. Tel.: +354 525 4818; fax: +354 552 8911;
e-mail: gghar@raunvis.hi.is
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.07.046

2894
A. Halldorsson et al. / Tetrahedron: Asymmetry 15 (2004) 2893–2899
site can be expected to readily accommodate the 1-O-
alkylglycerols. Although a large variety of diols, includ-
ing a whole range of 1-O-arylglycerols, have been
extensively studied in terms of kinetic resolution,13 there
are no reports on the 1-O-alkylglycerols of the type de-
scribed above. As far as we are aware the first and only
report dealing with this sort of substrates in lipase-cata-
lysed kinetic resolution is our communication published
in 1999.14
the first step, solketal was treated with grounded potas-
sium hydroxide and the corresponding alkyl bromides at
35–40ꢁC in the absence of solvent using tetrabutyl-
ammonium bromide as
a phase-transfer catalyst.
This afforded the crude 2,3-O-isopropylidene protected
1-O-alkylglycerol intermediates 2a–c, which were char-
acterised, but usually introduced directly to the second
step without further purification. In the second step,
the 2a–c intermediates were deprotected using catalytic
amount of p-toluenesulfonic acid in a 30% water–THF
solution under reflux conditions. The pure racemic
1-O-alkylglycerols 1a, b and c were afforded in 80%,
80% and 84% overall yields, respectively.
2. Results and discussion
The 1-O-alkyl-sn-glycerols are highly valuable com-
pounds that can be prepared from the unsaponifiable
fraction of various liver oils from elasmobranch fish.
However, as a result of the great variety of alkyl groups
in their ether linkage moiety they are by no means readily
accessible as pure compounds from natural sources.2 The
chief objective of this study was to develop a successful
lipase-catalysed process to accomplish the natural S-con-
figured chimyl 1a, batyl 1b and selachyl 1c alcohols.
OR
OH
OR
OH
OH
p-TsOH
KOH, RBr
n-Bu4NBr
O
O
O
O
H2O, THF
(RS)-2a-c
(RS)-1a-c
a (R = C16H33
b (R = C18H37
c (R = C18H35
)
)
)
Scheme 1.
Optically active 1-O-alkylglycerols can be readily pre-
pared from enantiopure C3-synthons possessing one
asymmetric centre, for example, solketal (2,2-dimethyl-
1,3-dioxolane-4-methanol). As an example, the (S)-solk-
etal is advantageously prepared from D-mannitol15 or by
Sharpless dihydroxylation16 whereas L-serine17 or ascor-
bic acid18 may be converted into the (R)-solketal. How-
ever, various attempts based on lipase-catalysed kinetic
resolution of racemic solketal and its derivatives have
been reported.19 Disappointingly, these processes dis-
play poor enantioselectivity due to the symmetrical
shape and small size of the solketal molecule. Despite
low enantioselectivity of lipases toward solketal its two
enantiomers have been successfully prepared enantio-
pure by lipase.20,21
The existence of all these compounds is manifested in
the literature.26 However, all the products were fully
1
characterized by high field H and 13C NMR and IR
spectroscopy as well as elemental analysis and/or high
resolution mass spectrometry.
2.2. Lipase-catalysed transesterification of 1-O-alkyl-
glycerols
The kinetic resolution of 1a–c was based on a sequential
lipase-catalysed transesterification using vinyl acetate as
an acylating agent. In such sequential alcoholysis reac-
tion, the racemic diols were acetylated regioselectively
at the primary hydroxyl group without significant enan-
tioselection by the lipase. Therefore, the second acetyl-
ation of the secondary hydroxyl group is responsible
for the enantioselection in the reaction, converting the
(S)-monoacetate much faster into the diacetate. The
greater enantiopreference for the (S)-1-O-alkyl-3-acetyl-
glycerol complies with the Kazlauskas empirical rule.
The overall enzymatic process is illustrated in Scheme 2.
Work highly relevant to that described in this paper has
been published by Theil et al., based on the sequential
acylation process originally introduced by Sih and co-
workers.22 They resolved various 1-O-arylglycerols by
lipase-catalysed sequential transesterification.23 All their
substrates agreed with KazlauskasÕs empirical rule for
predicting the stereochemical outcome of kinetic resolu-
tion of secondary alcohols by the Burkholderia cepacia
lipase (BCL, formerly Pseudomonas cepacia lipase).24
Their investigation took a new direction when a second
variety was introduced to the other primary alcoholic
group, varying in size and/or shape.13 It was difficult
to fit the results with the Kazlauskas rule, which predicts
high E-values on secondary alcohols when substituents
around the steric centre differ greatly in size but not
shape. Therefore, Theil et al. developed a new three-
dimensional active site model for BCL using computer
modelling, which revealed that the two hydrophobic
pockets of the lipase differ in shape, rather than size.25
OR
OR
OR
OR
Lipase
AcO
H
OAc
H
Lipase
AcO
OH
OH
OH
+
OH
OAc
OAc
(R)-3a-c
OAc
(RS)-1a-c
Scheme 2.
(S)-4a-c
(RS)-3a-c
In such kinetic resolution, it is necessary to use condi-
tions favouring rapid and irreversible reactions to
achieve high selectivity. Otherwise, the enantiomeric ex-
cess of the product and the remaining substrate will de-
crease progressively, as the extent of the reverse reaction
increases. Vinyl acetate was used to minimise the extent
of the reverse reaction by irreversible tautomerisation of
its enol-leaving moiety into volatile acetaldehyde.27
2.1. Synthesis of 1-O-alkylglycerols
Racemates of the 1-O-alkylglycerols 1a–c were prepared
in a two-step synthesis as demonstrated in Scheme 1. In

A. Halldorsson et al. / Tetrahedron: Asymmetry 15 (2004) 2893–2899
2895
Four different commercially available lipases were tested
under various transesterification conditions using vinyl
acetate as an acylating agent. They include the immobi-
lised Rhizomucor miehei lipase (RML, Lipozyme RM
IM) and Candida antarctica B lipase (CAL-B, Novozym
435), from Novozyme A/S in Denmark, and two bacte-
rial lipases from Amano Enzymes Ltd. in UK, Pseudo-
monas fluorescens (PFL, formerly Pseudomonas sp.
lipase) and BCL. The Burkholderia and Pseudomonas
lipases were observed to be far superior to RML and
CAL-B in terms of enantiomeric ratios offered. Moreo-
ver, the performance offered by PFL was significantly
better than BCL.
low temperature by CAL-B has been demonstrated on
glycerol to synthesise regioisomerically pure 1,3-diacyl-
glycerols.29 The C. antarctica lipase has also been
exploited to synthesise homogeneous triacylglycerols at
elevated temperature.30
OR
OR
OR
CAL, 4˚C
in CHCl3
OH
OH
OH
OAc
OAc
CAL, 65˚C
AcO
AcO
OAc
1a-c
4a-c
3a-c
Scheme 3.
In the transesterification reactions, the first acetylation
proceeded very rapidly with 80–100% monoacetates
3a–c being formed during the first reaction hour. The
fast conversion into the monoacetate indicates that
low or no enantioselection took place in the first acetyl-
ation as was firmly established. However, in the second
acetylation the reaction rate decreased dramatically
after approximately 50% conversion had been reached.
It was decided to look upon the monoacetate as the
racemic substrate, rather than the diol, since almost no
enantioselection took place in the first step. The conver-
sion was therefore based only on the second acetylation.
The sn-3 enantiomer was the better substrate in all the
cases and therefore the sn-1 enantiomer was obtained
in high enantiopurity, at higher than 50% conversion,
when the monoacetate was isolated from the reaction
mixture.
In order to determine the enantiomeric excess of both
the mono- and diacetate fractions they were separated
by a preparative TLC and then hydrolysed to the origi-
nal 1-O-alkylglycerol form. The diols were then reacted
with (R)-(À)-1-(1-naphthyhl)ethyl isocyanate (P95% ee)
to form their corresponding diastereomers, which were
separated on analytical HPLC using UV/vis detection
(280nm). Since the isocyanate used was contaminated
with 2.5% of the (S)-antipode simple correction calcula-
tions were made to compensate for that. Reference sam-
ples of homochiral (S)-1-O-alkylglycerols were prepared
to establish the procedure and the absolute configura-
tion of the compounds involved. In order to further con-
firm the results, MosherÕs esters31 were also synthesised
in some cases and analysed by 13C and 19F NMR
spectroscopy.
The conversion of the transesterification reactions was
used for the E-value calculations using the method of
Sih and Wu28 The mono-/diacetate ratio was determined
by GLC analysis after preparing trimethylsilyl deriva-
tives of the monoacetate in the reaction mixture. When
using FID detection response factors were needed to
correct the ratio due to different intensity given by the
two constituents. The conversion results were confirmed
Solvent studies on chimyl alcohol 1a revealed that the
highest E-values are obtained with PFL in chloroform
(18), better than in a 30% acetone in chloroform mixture
(16), diethyl ether (11) or diisopropyl ether (10). These
results are demonstrated in Table 1. However, the reac-
tion rate in chloroform was significantly slower than in
the other solvents, which demanded a higher enzyme
to substrate ratio.
1
by H NMR analysis as well as calculations from the
enantiomeric excess values.28
Table 2 displays the results of the sequential resolution
of the 1-O-alkylglycerols with PFL at room temperature
as well as 4ꢁC. Due to poor activity of the lipase toward
the monoacetate under the applied conditions high
enzyme–substrate ratio and extended reaction time was
needed. Despite rather low enantiomeric ratios for all
the alcohols 1a–c they were high enough to allow enan-
tiomerically pure material to be obtained for the
slower reacting (R)-monoacetate at approximately 60%
conversion for both the chimyl and batyl alcohols and
nearly so for the selachyl alcohol. The (R)-monoacetate
corresponds to the natural (S)-configuration of the
Reference samples of the mono- and diacetates were pre-
pared in excellent yields (ꢀ90%) using lipase and charac-
terised by high field 1H and 13C NMR spectroscopy. The
C. antarctica lipase was used to esterify the end-position
of racemic 1a–c exclusively using 1.0M vinyl acetate in
chloroform at low temperature (0–4ꢁC) and short reac-
tion time. The same enzyme was used to esterify both
hydroxyl groups using neat vinyl acetate, extended reac-
tion time and 65ꢁC. These reactions are illustrated in
Scheme 3. Such highly regioselective esterification at
Table 1. Kinetic resolution of racemic chimyl alcohol 1a by P. fluorescens lipase in various solvents at room temperaturea
Solvent
E/S
Conv. (%)
Time (h)
Ee 3 (%)
Ee 4 (%)
E
1:3 Acetone/CHCl3
CHCl3
Et2O
2.2
3.9
2.8
2.8
61
54
64
67
24
34
8
P95
86
94
6216
74
18
11
10
54
47
(i-Pr)2O
6
P95
a E/S, enzyme to substrate ratio (wt/wt); % Conv., conversion as determined by GLC; ee, enantiomeric excess as determined by HPLC on
diastereomers of isocyanates; E, these are average enantiomeric ratios of several reactions under identical conditions.

2896
A. Halldorsson et al. / Tetrahedron: Asymmetry 15 (2004) 2893–2899
Table 2. Kinetic resolution of racemic chimyl 1a, batyl 1b and selachyl 1c alcohols by P. fluorescens lipase in chloroforma
Substrate
Temp. (ꢁC)
Conv. (%)
Time (h)
64
90
63
87
61
9285
Ee 3 (%)
Ee 4 (%)
E
1a
2
4
2
4
2
4
0
0
0
7P5 95
94
56
80
17
32
55
51
56
1b
1c
7P2 95
89
71
5214
30
38
2
93
56
12
a See abbreviations in Table 1. The enzyme–substrate ratio was 3.5.
1-O-alkylglycerols as was established by the homochiral
reference samples.
gas chromatograph or (Beaconsfield, UK) using a 30-
m capillary column, DB-225 30N 0.25mm (J & W Sci-
entific, Folsom, CA) with hydrogen as a carrier gas at
a flow rate of 1.6mLminÀ1. Detection was with FID
and areal quantitation was made with an auxiliary auto-
matic integrator. Analytical HPLC was performed on a
constaMetric 3200 solvent delivery system with spectro-
Monitor 3200 variable wavelength detector (280nm)
both from LDC Analytical using a Nucleosil 50-5
column (Art. 721210) from Macherey-Nagel (Easton,
PA).
From the results displayed in Table 2 it is evident that
lowering the temperature had a dramatic effect on the
saturated chained alcohols 1a and 1b, but no such effect
was observed for 1c.
3. Conclusion
Racemic 1-O-alkylglycerols were prepared very effi-
ciently and their natural 1-O-alkyl-sn-glycerol enantio-
mers afforded by kinetic resolution using lipase-
catalysed transesterification. The best results were
obtained for the P. fluorescens lipase in chloroform for
which relatively low to moderate E-values were ob-
tained. This, however, enabled the preparation of the
saturated 1-O-alkylglycerols virtually enantiomerically
pure (P95% ee), whereas for the monounsaturated 1-
O-alkylglycerol slightly inferior results were obtained
(93% ee) at approximately 60% conversion as monoace-
tates. Comparison with pure enantiomers obtained from
(R)-solketal confirmed that these products all had the
same absolute configuration as the natural 1-O-alkyl-
sn-glycerols, that is, the S-configuration.
The bacterial lipases from P. fluorescens (Amano AK)
and P. cepacia (Amano PS) were supplied by Amano
Enzyme Europe Ltd. (Milton Keynes, England) and em-
ployed directly as powder, without any preadjustment or
optimisation of pH. The immobilised fungal lipases
from C. antarctica (Novozym 435) and Rhizomucor mie-
hei (Lipozyme RM IM) were supplied as a gift from
Novozyme A/S (Bagsvaerd, Denmark). Solketal (isopro-
pylideneglycerol, 98%), 1-bromohexadecane (97%), 1-
bromooctadecane (98%), 1-bromo-cis-9-octadecene
(99%), tetrabutylammonium bromide (99%), vinyl ace-
tate (>99%), (R)-(À)-1-(1-napthyl)ethyl isocyanate
(P95% ee), and hexamethyldisilazane were purchased
from Sigma Chemicals (St. Louis, Missouri). The
Sodium chloride, sodium hydroxide, potassium hydrox-
ide, anhydrous magnesium sulphate, chloromethylsi-
lane, triethylamine, phosphoric acid, p-toluenesulfonic
acid monohydrate and preparative TLC plates (Art.
5721) were obtained from Merck (Darmstadt,
Germany). All solvents (n-hexane, petroleum ether
(40–65ꢁC), diethyl ether, diisopropyl ether, chloroform,
methanol, tetrahydrofuran, ethyl acetate, acetone, tolu-
ene, phenylchloride and pyridine) of analytical grade
were obtained from Acros Organics (Geel, Belgium)
and used without further purification unless otherwise
stated.
4. Experimental
4.1. General
1H, 13C and 19F nuclear magnetic resonance (NMR)
spectra were recorded on a Bruker AC 250 NMR spec-
trometer (Karlsruhe, Germany) in deuterated chloro-
form as a solvent. The number of carbon nuclei
behind each 13C signal is indicated in parentheses after
each chemical shift value, when there is more than one
carbon responsible for the peak. Infrared (IR) spectra
were collected on a Nicolet Avatar 360 FT-IR (E.S.P.)
Spectrophotometer (Madison, WI) on a KBr pellet.
The high resolution mass spectra (HRMS) were ac-
quired on a Micromass Q-Tof II mass spectrometer
equipped with an Z-spray atmospheric pressure ionisa-
tion chamber. The mass spectrometry parameters were
as described by Mu et al.32 All data analysis was done
on Micromass MassLynx software. Elemental analyses
were performed on Carlo Erba Strumentazione Elemen-
tal Analyzer (model 1106; Milan, Italy). Melting points
4.1.1. Synthesis of 1-O-hexadecylglycerol 1a. Isoprop-
ylideneglycerol (432mg, 3.27mmol), 1-bromohexade-
cane (998mg, 3.27mmol) and tetrabutylammonium
bromide (211mg, 0.65mmol) were placed in a 10ml
round-bottomed flask and the resulting mixture stirred
on a magnetic stirrer hotplate for 10min at room tem-
perature. Grounded potassium hydroxide (367mg,
6.54mmol) was slowly and carefully added and the mix-
ture stirred for 15h at 35–40ꢁC in an oil bath. Addi-
tional potassium hydroxide (92mg, 1.64mmol) was
added and the solution stirred for extra 3h at the same
temperature. The product was extracted into n-hexane
(40mL) and the organic phase washed three times with
were determined on a Buchi 520 melting point apparatus
¨
and are uncorrected. Analytical gas–liquid chromato-
graphy (GLC) was conducted on a Perkin–Elmer 8140

A. Halldorsson et al. / Tetrahedron: Asymmetry 15 (2004) 2893–2899
2897
water (10mL) and brine solution (10mL) before remov-
ing the solvent in vacuo on a rotary evaporator to afford
yellowish oil mainly containing the 1-O-hexadecyl iso-
propylideneglycerol intermediate 2a.
of 1-O-cis-9-octadecenylglycerol (869mg, 2.54mmol) in
84% yield that soon became liquid at ambient tempera-
ture. H NMR d 5.37–5.33 (m, 2H, @CHCH2), 3.89–
1
3.82(m, 1H,
J = 3.8Hz, 1H, CH2OH), 3.62(dd,
C
HOH), 3.71 (dd, J = 11.4Hz,
J = 11.4Hz,
The oil was placed in a 25mL round-bottomed flask to-
gether with p-toluenesulfonic acid (80mg), tetrahydrofu-
ran (12mL) and water (5mL) and the resulting solution
refluxed overnight. After two extractions with diethyl
ether (10mL), washing with water (15mL) and brine
solution (10mL), drying over anhydrous magnesium
sulphate and solvent removal in vacuo on a rotary evap-
orator white crystalline material was afforded. Recrys-
tallisation from n-hexane in a freezer (À18ꢁC) resulted
in highly pure crystals of the 1-O-hexadecylglycerol
(829mg, 2.62mmol) in 80% yield. Mp = 64.0–64.6ꢁC.
1H NMR d 3.90–3.82(m, 1H, C HOH), 3.72(dd, J =
11.4Hz, J = 3.9Hz, 1H, CH2OH), 3.64 (dd, J =
11.5Hz, J = 5.1Hz, 1H, CH2OH), 3.54 (dd, J = 9.7Hz,
J = 4.0Hz, 1H, CH2OCH2CH2), 3.49 (dd, J = 9.7Hz,
J = 5.8Hz, 1H, CH2OCH2CH2), 3.46 (2 · t, J =
6.6Hz, 2H, OCH2CH2), 2.39 (br s, 2H, OH), 1.57 (br
quintet, J = 6.7Hz, 2H, OCH2CH2), 1.35–1.18 (m,
26H, CH2) and 0.88 (br t, J = 6.3Hz, 3H, CH3) ppm.
13C d 72.5, 71.8, 70.4, 64.3, 31.9, 29.7 (5), 29.6 (4),
29.4 (2), 26.1, 22.7 and 14.1ppm. IR mmax 3300–3600
(br, O–H), 2919 (vs, C–H), 2849 (vs, C-H) cmÀ1. HRMS
(API): calcd for C19H40O3+H m/z 317.3056, found
317.3095 amu. Elemental analysis. Found: C, 72.04; H,
12.73. C19H40O3 requires C, 72.10; H, 12.74.
J = 5.3Hz, 1H, CH2OH), 3.52(dd, J = 9.7Hz, J =
4.1Hz, 1H, CH2OCH2CH2), 3.47 (dd, J = 9.7Hz,
J = 5.9Hz, 1H, CH2OCH2CH2), 3.50–3.42(m, H2,
OCH2CH2), 2.61 (br s, 2H, OH), 2.03–1.97 (m, 4H,
@CHCH2), 1.58–1.54 (quintet (br), J = 6.7Hz, 2H,
OCH2CH2), 1.37–1.16 (m, 22H, CH2), 0.87 (br t,
J = 6.6Hz, 3H, CH3) ppm. 13C d 129.9, 129.8, 72.4,
71.8, 70.5, 64.2, 31.9, 29.7, 29.5 (3), 29.4, 29.3 (3),
29.2, 27.2 (2), 26.0, 22.7 and 14.1ppm. IR mmax 3300–
3600 (br, O-H), 2922 (vs, C-H), 2853 (vs, C–H) cmÀ1
.
HRMS (API): calcd for C21H42O3+H m/z 343.3210,
found 343.3212amu.
4.1.4. Synthesis of 1-O-hexadecyl-3-acetylglycerol
3a. Immobilised C. antarctica lipase (100mg) was
added to a 10mL round-bottomed flask at 0–4ꢁC (ice-
bath) containing a mixture of 1-O-hexadecylglycerol
(502mg, 1.59mmol) and 1.0M vinyl acetate in chloro-
form (10mL, 10.0mmol). The resulting mixture was
gently stirred at 0–4ꢁC for 2h. Then, the lipase was re-
moved by filtration and the solvent removed in vacuo
on a rotary evaporator to afford pure 1-O-hexadecyl-3-
acetylglycerol (506mg, 1.41mmol) as a clear oil in 89%
yields. 1H NMR d 4.18 (dd, J = 11.4Hz, J = 4.6Hz,
1H, CH2OCO), 4.11 (dd, J = 11.4Hz, J = 5.9Hz, 1H,
CH2OCO), 4.04–3.95 (m, 1H, CHOH), 3.50 (dd,
J = 9.6Hz, J = 4.2Hz, 1H, CH2OCH2CH2), 3.49–3.42
(m, 2H, CH2OCH2CH2), 3.40 (dd, J = 9.6Hz,
J = 6.2Hz, 1H, CH2OCH2CH2), 2.53 (br s, 1H,
CHOH), 2.10 (s, 3H, OCOCH3), 1.57 (quintet (br),
J = 6.5Hz, 2H, OCH2CH2), 1.25 (m, 26H, CH2) and
0.88 (br t, J = 6.6Hz, 6H, CH3) ppm. 13C d 171.1,
71.8, 71.3, 68.8, 65.7, 31.9, 29.7 (5), 29.6 (4), 29.4,
29.3, 29.0, 26.1, 22.7, 20.9 and 14.1ppm.
4.1.2. Synthesis of 1-O-octadecylglycerol 1b. The same
procedure was followed as described for 1a above using
isopropylidene glycerol (433mg, 3.27mmol), 1-bromo-
octadecane (1090mg, 3.27mmol), tetrabutylammonium
bromide (211mg, 0.65mmol) and potassium hydroxide
(459mg, 8.18mmol). Recrystallisation from n-hexane
in a freezer (À18ꢁC) afforded highly pure crystals of
the 1-O-octadecylglycerol (903mg, 2.62mmol) in 80%
yield. Mp = 69.2–69.8ꢁC. 1H NMR d 3.90–3.82(m,
1H, CHOH), 3.72(dd, J = 11.4Hz, J = 3.8Hz, 1H,
4.1.5. Synthesis of 1-O-octadecyl-3-acetylglycerol
3b. The same procedure was followed as described
for 3a above using 1-O-octadecylglycerol (500mg,
1.45mmol), 1.0 M vinyl acetate in chloroform (10mL,
10.0mmol) and immobilised C. antarctica lipase
(100mg). Pure 1-O-octadecyl-3-acetylglycerol (511mg,
1.32mmol) was obtained as a clear oil in 91% yield.
1H NMR d 4.18 (dd, J = 11.4Hz, J = 4.6Hz, 1H, CH2O-
CO), 4.11 (dd, J = 11.4Hz, J = 5.9Hz, 1H, –CH2OCO),
4.04–3.96 (m, 1H, CHOH), 3.50 (dd, J = 9.6Hz,
J = 4.2Hz, 1H, CH2OCH2CH2), 3.49–3.42(m, H2,
CH2OH),
3.64
(dd,
J = 11.4Hz,
J = 5.2Hz,
1H, CH2OH), 3.53 (dd, J = 9.7Hz, J = 4.1Hz, 1H,
CH2OCH2CH2), 3.48 (dd, J = 9.7Hz, J = 5.9Hz, 1H,
CH2OCH2CH2), 3.46 (2 · t, J = 6.6Hz, 2H, OCH2CH2),
2.50 (br s, 2H, OH), 1.57 (quintet (br), J = 6.7Hz, 2H,
OCH2 CH2), 1.36–1.16 (m, 30H, CH2) and 0.87 (br t,
J = 6.3Hz, 3H, CH3) ppm.
13
C d 72.5, 71.9, 70.4,
64.3, 31.9, 29.7 (7), 29.6 (4), 29.5, 29.4, 26.1, 22.7 and
14.1ppm. IR mmax 3300–3600 (br, O–H), 2918 (vs, C–
H), 2849 (vs, C–H) cmÀ1. HRMS (API): calcd for
C21H44O3+H m/z 345.3369, found 345.3333 amu. Ele-
mental analysis. Found: C, 73.08; H, 12.91. C21H44O3
requires C, 73.20; H, 12.87.
OCH2CH2), 3.42(dd,
J = 9.6Hz, J = 6.1Hz, 1H,
CH2OCH2CH2), 2.53 (br s, 1H, CHOH), 2.10 (s, 3H,
OCOCH3), 1.57 (quintet (br), J = 6.7Hz, 2H,
OCH2CH2), 1.28–1.22 (m, 30H, CH2) and 0.88 (br t,
J = 6.6Hz, 6H, CH3) ppm. 13C d 171.2, 71.8, 71.3,
68.8, 65.7, 31.9, 29.7 (7), 29.6 (4), 29.5, 29.3, 26.0, 22.7
and 14.1ppm.
4.1.3.
Synthesis
of
1-O-cis-9-octadecenylglycerol
1c. The same procedure was followed as described
for 1a above using isopropylidene glycerol (399mg,
3.02mmol),
1-bromo-cis-9-octadecene
(1001mg,
3.02mmol), tetrabutylammonium bromide (195mg,
0.60mmol) and potassium hydroxide (418mg,
7.45mmol). Recrystallisation from n-hexane in chilled
chlorobenzene (À42ꢁC) afforded highly pure crystals
4.1.6. Synthesis of 1-O-cis-9-octadecenyl-3-acetylglycerol
3c. The same procedure was followed as described for
3a above using 1-O-cis-9-octadecenylglycerol (497mg,
1.45mmol), 1.0M vinyl acetate in chloroform (10mL,

2898
A. Halldorsson et al. / Tetrahedron: Asymmetry 15 (2004) 2893–2899
10.0mmol) and immobilised C. antarctica lipase
(100mg). Pure 1-O-cis-9-octadecyl-3-acetylglycerol
(505mg, 1.31mmol) was obtained as a clear oil in 90%
J = 3.7Hz, 1H, CH2OCO), 4.16 (dd, J = 12.0Hz,
J = 6.4Hz, 1H, CH2OCO), 3.54 (d, J = 5.3Hz,
CH2OCH2CH2), 3.48–3.39 (m, 2H, OCH2 CH2), 2.09
(s, 3H, OCOCH3), 2.07 (s, 3H, OCOCH3), 2.07–1.98
(m, 4H, @CHCH2), 1.55 (quintet (br), J = 6.4Hz, 2H,
1
yield. H NMR d 5.37–5.33 (m, 2H, @CHCH2), 4.18
(dd, J = 11.4Hz, J = 4.5Hz, 1H, CH2OCO), 4.10 (dd,
J = 11.4Hz, J = 5.8Hz, 1H, CH2OCO), 4.03–3.97 (m,
1H, CHOH), 3.50 (dd, J = 9.6Hz, J = 4.2Hz, 1H,
CH2OCH2CH2), 3.49–3.43 (m, 2H, OCH2CH2), 3.42
(dd, J = 9.6Hz, J = 6.1Hz, 1H, CH2OCH2CH2), 2.53
(br s, 1H, CHOH), 2.10 (s, 3H, OCOCH3), 2.05–1.98
(m, 4H, @CHCH2), 1.57 (quintet (br), J = 6.5Hz, 2H,
OCH2CH2), 1.29–1.27 (m, 22H, CH2) and 0.88 (br t,
J = 6.6Hz, 6H, –CH3) ppm. 13C d 171.1, 129.9, 129.8,
71.8, 71.3, 68.8, 65.7, 31.9, 29.8, 29.6, 29.5 (2), 29.4,
29.3 (2), 29.2 (2), 27.2 (2), 26.1, 22.7, 20.9 and 14.1ppm.
OCH2CH2), 1.29–1.26 (br s, 22H, CH2) and 0.88 (t,
13
J = 6.6Hz, 3H, CH3) ppm.
C d 170.7, 170.4, 129.9,
129.8, 71.7, 70.3, 68.8, 63.0, 31.9, 29.7, 29.5 (3), 29.4,
29.3 (4), 27.2 (2), 26.0, 22.7, 22.7, 21.0, 20.8 and 14.1ppm.
4.2. Lipase-catalysed transesterification of 1-O-
alkylglycerols
In a typical procedure the 1-O-alkylglycerol (50mg) was
dissolved in 1.0M vinyl acetate in chloroform (2mL) in
a 10mL round-bottom flask before adding the powdered
P. fluorescens lipase (200mg). To monitor the progress
of the reaction small sample (approximately 100lL)
was taken, the enzyme removed by filtration and the
solvent removed in vacuo on a rotary evaporator. To
the sample pyridine (500lL), hexamethyldisilazane
(150lL) and chloromethylsilane (50lL) were added
and the resulting mixture shaken vigorously for 1min
and left standing for three minutes. After centrifugation
the clear solution of TMS derivatives was injected into a
GLC. A temperature of 185ꢁC was used at the begin-
ning, rising 3ꢁCminÀ1 to a final temperature of 210ꢁC,
which was maintained for 10–23min (depending on
the alkyl chain). Injector and detector temperatures were
kept at 265ꢁC. The GLC results were corrected by re-
sponse factors due to the FID detection and the factors
1.01 and 1.29 were used for 3a–c and 4a–c, respectively.
The conversion was also determined by 1H NMR
spectroscopy.
4.1.7. Synthesis of 1-O-hexadecyl-2,3-diacetylglycerol
4a. Immobilised C. antarctica lipase (100mg) was
added to a 10mL round-bottomed flask containing a
mixture of 1-O-hexadecylglycerol (500mg, 1.58mmol)
and vinyl acetate (2.01g, 23.2mmol). The resulting mix-
ture was gently stirred at 65ꢁC for 17h under nitrogen
atmosphere. Then, the lipase was removed by filtration
and the excess vinyl acetate removed in vacuo on a vac-
uum pump (0.01Torr) to afford pure 1-O-hexadecyl-2,3-
diacetylglycerol (567mg, 1.41mmol) as a clear oil in
89% yield. 1H NMR d 5.21–5.13 (m, 2H, CHOAc),
4.32(dd,
J = 12.0Hz, J = 3.7Hz, 1H, CH2OCO),
4.14 (dd, J = 12.0Hz, J = 6.4Hz, 1H, CH2OCO), 3.53
(d, J = 5.3Hz, CH2OCH2CH2), 3.46–3.37 (m, 2H,
OCH2CH2), 2.07 (s, 3H, OCOCH3), 2.05 (s, 3H,
OCOCH3), 1.53 (quintet (br), J = 6.6Hz, 2H,
OCH2CH2), 1.24 (br s, 26H, CH2) and 0.86 (t,
J = 6.6Hz, 3H, CH3) ppm. 13C d 170.6, 170.3, 71.7,
70.3, 68.8, 62.9, 31.9, 29.7 (5), 29.6 (3), 29.5, 29.4,
29.3, 26.0, 22.6, 21.0, 20.7 and 14.1ppm.
4.2.1. Separation and hydrolysis of the reaction mix-
ture. After desired conversion was obtained the lipase
was removed by filtration and the solvent evaporated.
The resulting mixture was separated on a preparative
TLC-plate using 4% acetone in chloroform as eluent.
The products 3a–c (Rf 0.15–0.25) and 4a–c (Rf 0.75–
0.90) were scraped off and extracted from the silica gel
with ethyl acetate. After removal of the silica gel by fil-
tration and evaporation of the solvent the two products
were hydrolysed to afford their corresponding diols (1a–
c). Each product was dissolved in 0.5M sodium hydrox-
ide in methanol (2mL) and reacted for 30min at 60 ꢁC.
After cooling 0.5M phosphoric acid (2mL) and water
(2mL) were added and the resulting mixture extracted
twice with diethyl ether. The organic layer was dried
over anhydrous magnesium sulphate and the solvent
removed in vacuo on a rotary evaporator. The resulting
diols were derivatised with enantiopure (R)-(À)-
1-(1-naphthyl)ethyl isocyanate and in some cases
(R)-(+)-a-methoxy-a-(trifluoromethyl)phenylacetic acid31
(Mosher acid) for comparison.
4.1.8. Synthesis of 1-O-octadecyl-2,3-diacetylglycerol
4b. The same procedure was followed as described
for 4a above using 1-O-octadecylglycerol (503mg,
1.46mmol), vinyl acetate (2.04g, 23.7mmol) and immo-
bilised C. antarctica lipase (100mg). Pure 1-O-octadecyl-
2,3-diacetylglycerol (564mg, 1.32mmol) was obtained as
1
a clear oil in 90% yield. H NMR d 5.20–5.16 (m, 2H,
CHOAc), 4.33 (dd, J = 12.0Hz, J = 3.7Hz, 1H, CH2O-
CO), 4.16 (dd, J = 12.0Hz, J = 6.3Hz, 1H, CH2OCO),
3.54 (d, J = 5.3Hz, CH2OCH2CH2), 3.47–3.40 (m, 2H,
OCH2CH2), 2.08 (s, 3H, OCOCH3), 2.06 (s, 3H,
OCOCH3), 1.54 (quintet (br), J = 6.6Hz, 2H,
OCH2CH2), 1.25 (br s, 30H, CH2) and 0.88 (t,
13
J = 6.6Hz, 3H, CH3) ppm.
70.3, 68.8, 62.9, 31.9, 29.6 (10), 29.5, 29.4, 29.3, 26.0,
C d 170.6, 170.3, 71.7,
22.6, 21.0, 20.7 and 14.1ppm.
4.1.9. Synthesis of 1-O-cis-9-octadecenylglycerol-2,3-
diacetylglyceol 4c. The same procedure was followed
as described for 4a above using 1-O-cis-9-octadecylgly-
cerol (498mg, 1.45mmol), vinyl acetate (2.01g,
23.3mmol) and immobilised C. antarctica lipase
(100mg). Pure 1-O-cis-9-octadecenyl-2,3-diacetylgly-
cerol (552mg, 1.29mmol) was obtained as a clear oil
4.2.2. Preparation of (R)-(À)-1-(1-naphthyl)ethyl isocya-
nate derivatives. The corresponding diols of 3a–c and
4a–c (1–2mg) obtained above were mixed with 2% tri-
ethylamine in toluene (1mL) and (R)-(À)-1-(1-naph-
thyl)ethyl isocyanate (3lL). The resulting mixture was
stirred at room temperature under nitrogen over night.
1
in 89% yield. H NMR d 5.37–5.32(m, 2H, @CHCH2),
5.23–5.15 (m, 2H, CHOAc), 4.34 (dd, J = 12.0Hz,

A. Halldorsson et al. / Tetrahedron: Asymmetry 15 (2004) 2893–2899
2899
11. Bornscheuer, U. T.; Kazlauskas, R. J. Hydrolases in
Organic Synthesis––Regio- and Stereoselective Biotransfor-
mations; Wiley-VCH: Weinheim, 1999, pp 65–129.
12. Haraldsson, G. G. In The Chemistry of the Functional
Groups, Supplement B2: The Chemistry of Acid Deriva-
tives; Patai, S., Ed.; John Wiley & Sons: Chichester, 1992,
pp 1395–1473.
13. Theil, F.; Lemke, K.; Ballschuh, S.; Kunath, A.; Schick,
H. Tetrahedron: Asymmetry 1995, 6, 1323–1344.
14. Haraldsson, G. G.; Thordarson, P.; Halldorsson, A.;
Kristinsson, B. Tetrahedron: Asymmetry 1999, 10,
3671–3674.
The isocyanate derivatives 3a–c were injected directly to
analytical HPLC using a 1:1 n-hexane/ethyl acetate as
eluent flowing at 0.15mlminÀ1. The diastereomers cor-
responding to the (R)-3-O-alkyl-sn-glycerol (24 min
retention time) and the (S)-1-O-alkyl-sn-glycerol
(27min retention time) were successfully separated and
the enantiomeric ratio calculated from their peak area.
Approximately 2.5% (S)-(À)-1-(1-naphthyl)ethyl isocya-
nate impurities in the isocyante used had to be corrected
for in the enantiomeric ratio determination.
15. Csuk, R.; Glanzer, B. I. Chem. Rev. 1991, 91, 49–97.
16. Wang, Z. M.; Sharpless, K. B. Tetrahedron Lett. 1993, 34,
8225–8228.
Acknowledgements
17. Lok, C. M.; Ward, J. P.; Van Drop, D. A. Chem. Phys.
Lipids 1976, 16, 115–122.
18. Jung, M. E.; Shaw, T. J. J. Am. Chem. Soc. 1980, 102,
6304–6311.
19. Phung, A. N.; Braun, J.; Le Goffic, F. Synth. Commun.
1995, 25, 1783–1788.
20. Va¨nttinen, E.; Kanerva, L. T. J. Chem. Soc., Perkin Trans.
1 1994, 3459–3463.
The University of Iceland Research Fund is acknowl-
edged for financial support, Dr. Sigridur Jonsdottir at
the Science Institute for running the NMR analysis, Dr.
Sigurdur V. Smarason at deCode Genetics for perform-
ing the mass spectrometry analysis, and Svana Stefans-
dottir at the University of Iceland for elemental analysis.
21. Va¨nttinen, E.; Kanerva, L. T. Tetrahedron: Asymmetry
1997, 8, 923–933.
References
22. Guo, Z.-W.; Wu, S.-H.; Chen, C.-S.; Girdaukas, G.; Sih,
C. J. J. Am. Chem. Soc. 1990, 112, 4942–4974.
23. Theil, F.; Weidner, J.; Ballschuh, S.; Kunath, A.; Schick,
H. J. Org. Chem. 1994, 59, 388–393.
24. Kazlauskas, R. J.; Weissfloch, A. N. E.; Rappaport, A. T.;
Cuccia, L. A. J. Org. Chem. 1991, 56, 2656–2665.
25. Lemke, K.; Lemke, M.; Theil, F. J. Org. Chem. 1997, 62,
6268–6273.
26. Dictionary of Organic Compounds; Chapman and Hall:
New York, 1982; Vol. 2, pp 1945, and 1990.
27. Wang, Y.-F.; Lalonde, J. J.; Momongan, M.; Berbreiter,
D. E.; Wong, C.-H. J. Am. Chem. Soc. 1988, 110,
7200–7205.
28. Sih, C. J.; Wu, S.-H.; Eliel, E. L.; Wilen, S. H.; Eds.;
Topics in Stereochemistry; John Wiley & Sons: New York,
1989; Vol. 19, pp 63–126.
1. Ether Lipids. Biochemical and Biomedical Aspects; Man-
gold, H. K., Palthauf, F., Eds.; Academic: New York,
1983.
2. Kayama, M.; Mankura, M. Inform 1998, 9, 794–799.
3. Sargent, J. R. In Marine Biogenic Lipids, Fats and Oils;
Ackman, R. G., Ed.; CRC: Boca Raton, FL, 1989, Vol. 1;
pp 175–197.
4. Benveniste, J.; Vargaftig, B. B. In Ether Lipids––Biochem-
ical and Biomedical Aspects; Mangold, H. K., Palthauf, F.,
Eds.; Academic: New York, 1983; pp 356–373.
5. Mangold, H. K.; Weber, N. Lipids 1987, 22, 789–799.
6. Brohult, A.; Brohult, J.; Brohult, S. Acta Chem. Scand.
1970, 24, 730–733.
7. Pugliese, P. T.; Jordan, K.; Cederberg, H.; Brohult, J.
J. Alt. Compl. Med. 1998, 4, 87–99.
8. Christie, W. W. Lipid Analysis––Isolation, Separation,
Identification and Structural Analysis of Lipids; The Oily
Press: Bridgwater, 2003, pp 3–33.
29. Halldorsson, A.; Magnusson, C. D.; Haraldsson, G. G.
Tetrahedron 2003, 59, 9101–9109.
¨
¨
30. Haraldsson, G. G.; Gudmundsson, B. O.; Almarsson, O.
Tetrahedron 1995, 51, 941–952.
31. Dale, J. A.; Mosher, S. H. J. Am. Chem. Soc. 1973, 95,
512–519.
32. Mu, H.; Sillen, H.; Høy, C.-E. J. Am. Oil Chem. Soc. 2000,
77, 1049–1059.
9. Wong, C. H.; Whitesides, G. M. Enzymes in Synthetic
Organic Chemistry; Pergamon: Oxford, 1994, pp 70–111.
10. Faber, K. Biotransformations in Organic Synthesis;
Springer: Heidelberg, 1995, pp 88–115.