Highly regioselective propanoylation of dihydroxybenzenes mediated by Candida antarctica lipase B in organic solvents

Article abstract of DOI:10.1016/j.tetlet.2007.10.124

Candida antarctica lipase B (CAL-B) was found to be a highly active biocatalyst for the direct acylation of the phenolic hydroxyls of substituted hydroquinones and resorcinols with vinyl propanoate as an acyl donor. The acylation reactions took place gene

Full text of DOI:10.1016/j.tetlet.2007.10.124

Available online at www.sciencedirect.com
Tetrahedron Letters 49 (2008) 175–178
Highly regioselective propanoylation of dihydroxybenzenes
mediated by Candida antarctica lipase B in organic solvents
Toshifumi Miyazawa,* Manabu Hamada, Ryohei Morimoto, Takashi Murashima
and Takashi Yamada
Department of Chemistry, Faculty of Science and Engineering, Konan University, Higashinada-ku, Kobe 658-8501, Japan
Received 19 September 2007; revised 22 October 2007; accepted 25 October 2007
Available online 30 October 2007
Abstract—Candida antarctica lipase B (CAL-B) was found to be a highly active biocatalyst for the direct acylation of the phenolic
hydroxyls of substituted hydroquinones and resorcinols with vinyl propanoate as an acyl donor. The acylation reactions took place
generally in a very regioselective manner. Especially in the case of 4-substituted resorcinols, the hydroxyl remote from the substitu-
ent was regiospecifically acylated to afford only the 1-O-propanoylated resorcinols.
Ó 2007 Elsevier Ltd. All rights reserved.
The high potential of lipases (triacylglycerol hydrolases,
E.C. 3.1.1.3) as biocatalysts for organic synthesis has
been well documented.1 Lipases from a variety of
sources have been utilized for preparing homochiral
compounds related to pharmaceuticals and agrochemi-
cals through resolution or desymmetrization. Besides
the stereoselective nature of lipases, their regioselective
properties have also been exploited for the preparation
of selectively protected derivatives of compounds con-
taining multi-hydroxyl groups such as carbohydrates
and steroids. Thus, the lipase-catalyzed acylation or
deacylation procedure has proven to be much superior
to the standard chemical methodology which often re-
quires a series of tedious steps for such transformations.
These enzymatic acyl-transfer approaches can protect or
deprotect a hydroxyl in the presence of several others
under optimized reaction conditions. Compared to such
studies on alcoholic hydroxyls, much fewer examples
have been accumulated on phenolic hydroxyls. Further
attention should be paid to the exploitation of lipase’s
ability to discriminate between hydroxyls of this type,
which usually show only a small difference in reactivity
toward conventional chemical reagents. Although sev-
eral papers have dealt with the lipase-catalyzed deacyla-
tion of peracetylated polyhydroxybenzenes2 and
flavonoids3 via transesterification with an alcohol such
as 1-butanol,4 the lipase-catalyzed direct acylation of
phenolic hydroxyls have rarely been reported. This is
probably because a phenolic hydroxyl is generally far
less nucleophilic than an alcoholic hydroxyl and/or phe-
nolic compounds are known to inhibit some enzymes.
Nicolosi and co-workers reported the lipase-catalyzed
regioselective protection of hydroxy groups in aromatic
dihydroxyaldehydes and ketones:5,6 the hydroxyl other
than the one at position ortho to the carbonyl was selec-
tively acylated employing Burkholderia (Pseudomonas)
cepacia lipase and vinyl acetate in cyclohexane–t-amyl
alcohol (9:1). They attributed the high regioselectivity
observed to the chelation of the carbonyl with the o-hy-
droxy group. Stimulated by this work, we have started
to examine the regioselectivity in the lipase-catalyzed
acylation of the simplest members of polyphenols, that
is, dihydroxybenzenes (hydroquinones and resorcinols)
carrying substituents other than the carbonyl. The prep-
aration of regioselectively protected derivatives of poly-
phenolic compounds by direct acylation procedure
seems to be an extremely challenging task.
Initially, several hydroquinones bearing different sub-
stituents including the acetyl group on the benzene ring
(1) were subjected to acylation with vinyl propanoate
(3 M equiv) in the presence of immobilized B. cepacia
lipase (Amano lipase PS) on Celite in the same mixed
solvent system as mentioned above at 45 °C (Scheme
1). Hydroquinone 1 can undergo enzymatic acylation
through two pathways to form either the 4-O-propanoyl
Keywords: Candida antarctica lipase B; Regioselective acylation;
Regioselectivity; Hydroquinones; Resorcinols.
*
Corresponding author. Tel.: +81 78 431 4341; fax: +81 78 435
2539; e-mail: miyazawa@konan-u.ac.jp
0040-4039/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2007.10.124

176
T. Miyazawa et al. / Tetrahedron Letters 49 (2008) 175–178
OH
OCOC2H5
R
R'
OH
R
OCOC2H5
OH
OCOC2H5
R
OCOC2H5
OH
R
6
2
R
R'
R'
OCOC2H5
OH
OH
OCOC2H5
R
OCOC2H5
R
OH
R
4
1
8
5
R'
OCOC2H5
OH
3
7
Scheme 1. CAL-B-catalyzed regioselective acylation of substituted
hydroquinones (1). R: a, Me; b, Me; c, Et; d, i-Pr; e, t-Bu; f, MeO; g,
Ac; h, F; i, Cl; j, Br. R0: a and c–j, H; b, Me.
Scheme 2. CAL-B-catalyzed regioselective acylation of 4-substituted
resorcinols (5). R: a, Me; b, Et; c, (CH3)3CCH2C(CH3)2–; d, Bn; e, Cl;
f, Br.
While the regioselectivity was very low with B. cepacia
lipase,10 the acylation catalyzed by CAL-B proceeded
regiospecifically: the 1-O-propanoyl derivative (6b) was
obtained as the sole product.11 These results indicate
that CAL-B12 was an enzyme of choice for the direct
acylation of phenolic hydroxyls. Accordingly, further
studies were conducted using this lipase. Although a
large number of examples have already been accumu-
lated on the effect of organic solvents on lipase-catalyzed
reactions,13 it is still difficult to predict the solvent effect,
especially on regioselectivity. Screening experiments for
an appropriate solvent revealed that CAL-B was most
active with a tolerable regioselectivity in diisopropyl
ether.
derivative (2) or the 1-O-propanoyl derivative (3), and
finally to afford the 1,4-di-O-propanoyl derivative (4).7
The results obtained after 3 days of incubation are
shown in Table 1. The phenolic hydroxyls of all the
hydroquinones (1) examined managed to be acylated,
though the reactions were rather slow. The reaction with
acetylhydroquinone (1g) was the slowest among those
with the substituted hydroquinones examined, and be-
sides the expected 4-O-propanoyl derivative (2g) the iso-
meric 1-O-propanoyl derivative (3g) was also produced
in a small amount, indicating the incompleteness of this
lipase’s regioselectivity even toward the aromatic
dihydroxyketone. However, it exhibited some regioselec-
tivity toward the hydroquinones carrying substituents
other than the carbonyl: the hydroxyl remote from the
substituent was preferentially acylated. In the case of
methylhydroquinone (1a) bearing the smallest substitu-
ent, the regioselectivity was very poor. Accordingly,
lipases from microbial and pancreatic sources8 were
screened for the acylation of 1a with vinyl propanoate
at 45 °C in diisopropyl ether.9 Of the enzymes tested,
Candida antarctica lipase B (CAL-B) proved to be the
best in terms of both activity and regioselectivity.
The screening of lipases was also conducted employing
4-ethylresorcinol (5b) and vinyl propanoate (Scheme
2). In general, the acylations of 5b were slower than
those of 1a.
Using immobilized CAL-B14 as the biocatalyst in diiso-
propyl ether the product distribution of a number of
hydroquinones (1) carrying different substituents on
the benzene ring was examined in the acylation with
vinyl propanoate.15 The time-course of the propanoyla-
tion of methoxyhydroquinone (1f) was followed through
the quantification of the products by HPLC analysis
(Fig. 1). The starting hydroquinone was consumed
abruptly during the first 6 h, then decreased gradually,
100
80
60
40
20
0
Table 1. B. cepacia lipase-catalyzed propanoylation of substituted
hydroquinones (1) with vinyl propanoatea
Substrate
R
R0
Yield (%)
2
3
4
1a
1b
1e
1f
CH3
CH3
(CH3)3C
CH3O
CH3CO
H
CH3
H
H
H
41.3
78.8
73.2
41.1
21.7
26.9
3.5
1.9
15.6
5.8
30.4
0
1.2
12.9
0
0
10
20
30
40
50
Time (h)
1g
a Reactions were conducted using 1 (0.1 mmol), vinyl propanoate
(0.3 mmol), and B. cepacia lipase immobilized on Celite (40 mg) in
240 ll of cyclohexane–t-amyl alcohol (9:1, v/v) at 45 °C for 3 days.
Figure 1. Reaction profile in the CAL-B-catalyzed acylation of
methoxyhydroquinone (1f) with vinyl propanoate in diisopropyl ether.
Symbols: diamond, 1f; circle, 2f; triangle, 3f; square, 4f.

T. Miyazawa et al. / Tetrahedron Letters 49 (2008) 175–178
177
Table 2. CAL-B-catalyzed propanoylation of substituted hydroqui-
nones (1) with vinyl propanoatea
Table 3. CAL-B-catalyzed propanoylation of substituted resorcinols
(5) with vinyl propanoatea
Substrate
R
R0
Yield (%)
Substrate
R
Yield (%) of 6b
2
3
4
5a
5b
5c
5d
5e
5f
CH3
CH3CH2
(CH3)3CCH2C(CH3)2
C6H5CH2
Cl
Br
82.2
37.0
23.1
36.7
74.1
69.0
1a
1c
1d
1e
1f
1h
1i
CH3
H
H
H
H
H
H
H
H
51.4
73.2
75.0
93.3
70.7
39.0
40.9
47.0
17.5
8.4
0
17.2
11.5
16.0
0
18.6
19.0
4.7
CH3CH2
(CH3)2CH
(CH3)3C
CH3O
F
0
10.5
33.0
16.4
22.1
a Reactions were conducted using 5 (0.1 mmol), vinyl propanoate
(0.3 mmol), and CAL-B (40 mg) in diisopropyl ether (240 ll) at 45 °C
for 3 days.
b The 3-O-propanoate (7) and 1,3-di-O-propanoate (8) were not
detected in all the cases.
Cl
Br
1j
7.7
a Reactions were conducted using 1 (0.1 mmol), vinyl propanoate
(0.3 mmol), and CAL-B (40 mg) in diisopropyl ether (240 ll) at 45 °C
for 1 day.
yielding the 1-O-propanoyl derivative (6) as the exclu-
sive product and none of the 3-O-propanoyl derivative
(7) nor the 1,4-di-O-propanoyl derivative (8). The steric
factor of substituents in the substituted resorcinols had a
much larger effect on its activity than that in the substi-
tuted hydroquinones. With 4-methylresorcinol (5a) the
acylation proceeded rather smoothly among the resor-
cinols examined. It was retarded in the presence of bulk-
ier substituents. This is probably because of the less
accessibility of the lipase even toward the hydroxyl at
position 1, to say nothing of the hydroxyl at position
3. The acylation of halogen-substituted resorcinols (5e
and 5f) proceeded at the rate between those of 5a and
the other alkyl-substituted resorcinols.
and finally disappeared after 24 h. The amount of
the 4-O-acylated product (2f) increased steeply at first,
then reached a maximum after ca. 12 h and then turned
to a gradual decrease. Thus, the product distribution
was dependent largely on the reaction time. Accord-
ingly, the results obtained after 1 day of incubation at
45 °C are compiled in Table 2. In all the cases, the 4-
O-propanoyl derivative (2) was obtained as the main
product, and the 1-O-propanoyl derivative (3) and/or
the 1,4-di-O-propanoyl derivative (4) were also pro-
duced in various amounts depending on the substitu-
ents. When the substituent was changed from a methyl
group (in 1a) to an ethyl group (in 1c), the proportion
of the 4-O-propanoyl derivative in the reaction products
increased to a considerable extent. With isopropylhy-
droquinone (1d) the formation of the 1-O-propanoyl
derivative (3d) was not observed, and the 4-O-propanoyl
derivative (2e) was produced as the exclusive product
with t-butylhydroquinone (1e). Thus, the selectivity for
the propanoylation of the hydroxyl remote from the
substituent increases with the increase in the bulk of
the substituent, that is, methyl to isopropyl/t-butyl. This
implies that the steric bulk of the substituent must be
one of the contributory factors for the observed regiose-
lectivity. The behavior of 1f resembles that of 1c, which
also suggests the importance of steric requirement in
activity and regioselectivity. With fluorohydroquinone
(1h) the regioselectivity deteriorated to a great extent,
though its activity toward the acylation was as high as
those of the alkyl-substituted hydroquinones. The other
halogen substituents (in 1i and 1j) reduced the activity of
hydroquinone, with a moderate regioselectivity. The
effect of halogen substituents on the activity and regio-
selectivity seems rather complicated, indicating that
besides the steric effect other factors, such as the elec-
tronic effect, must be taken into consideration.
With the purpose of comparison some typical non-enzy-
matic procedures were also tried for the acylation of
phenolic hydroxyls. When 4-ethylresorcinol (5b) was
heated in propanoic anhydride at 100 °C for 30 min,
the product distribution was as follows: 6b, 16.8%; 7b,
16.8%; 8b, 32.9%. In the presence of a few drops of sul-
furic acid the acylation with propanoic anhydride was
accelerated to a great extent. The reaction at 25 °C
yielded 20.0% of 6b, 0% of 7b, and 80.0% of 8b after
5 min.16 On the other hand, the reaction of 5b with pro-
panoyl chloride (1 M equiv) in toluene in the presence of
pyridine (1 M equiv) as catalyst at 50 °C gave 57.0% of
6b, 12.2% of 7b, and 30.8% of 8b after 30 min. Thus,
the acylations by the conventional methods were rather
faster than the CAL-B-catalyzed acylation, though the
difference in the reaction conditions makes a direct com-
parison difficult. However, it is certain that these chem-
ical procedures were much inferior in terms of
regioselectivity, though a preference for the 1-O-propa-
noyl derivative (6) was observed to one degree or
another in some cases.
In conclusion, CAL-B is a highly active biocatalyst for
the direct acylation of the phenolic hydroxyls of hydro-
quinones and resorcinols. The acylation reactions take
place generally in a very regioselective manner; in partic-
ular, they are regiospecific in the case of resorcinols. The
main or, in some cases, exclusive products obtained
through the direct acylation of those dihydroxybenzenes
are the regioisomers of those obtained through the
CAL-B-catalyzed deacylation of dihydroxybenzenes
acylated at both phenolic hydroxyls.4 This should be
Next, the CAL-B-catalyzed propanoylation of a number
of resorcinols (5) carrying alkyl, aralkyl, or halogen sub-
stituents on the benzene ring was examined under the
same reaction conditions as above. The results obtained
after 3 days of incubation at 45 °C are shown in Table 3.
The reactions were retarded largely compared with those
of hydroquinones having the same or similar substitu-
ents. Quite interestingly, however, the acylation pro-
ceeded regiospecifically independent of the substituent,

178
T. Miyazawa et al. / Tetrahedron Letters 49 (2008) 175–178
and C-6 appeared in the HMBC spectrum of 2e. Selected
data for 2e: oil; 1H NMR (DMSO-d6) d 1.11 (3H, t,
J = 7.5 Hz, CH3), 1.32 (9H, s, C(CH3)3), 2.53 (2H, q,
J = 7.5 Hz, CH2), 6.73-6.75 (1H, distorted dd, J = 8.5 and
2.0 Hz, H-5), 6.75-6.77 (1H, distorted d, J = 8.5 Hz, H-6),
6.80 (1H, d, J = 2.0 Hz, H-3), 9.39 (1H, s, OH). For 3e:
oil; 1H NMR (DMSO-d6) d 1.14 (3H, t, J = 7.5 Hz,
CH2CH3), 1.24 (9H, s, C(CH3)3), 2.58 (2H, q, J = 7.5 Hz,
CH2CH3), 6.58 (1H, dd, J = 3.0 and 8.5 Hz, H-5), 6.74
(1H, d, J = 3.0 Hz, H-3), 6.77 (1H, d, J = 8.5 Hz, H-6),
9.29 (1H, s, OH). On the other hand, the authentic
samples of the dipropanoates of dihydroxybenzenes were
prepared by the reaction of each dihydroxybenzene with
propanoyl chloride in pyridine. Selected data for 4e:
oil; 1H NMR (DMSO-d6) d 1.13 (3H, t, J = 7.5 Hz,
CH2CH3), 1.17 (3H, t, J = 7.5 Hz, CH2CH3), 1.28 (9H, s,
C(CH3)3), 2.59 (2H, q, J = 7.5 Hz, CH2), 2.65 (2H, q, J =
7.5 Hz, CH2), 7.01 (1H, dd, J = 9.0 and 2.5 Hz, H-5), 7.06
(1H, d, J = 2.5 Hz, H-3), 7.07 (1H, d, J = 9.0 Hz, H-6).
8. The lipases tested include those from Candida antarctica
B, Alcaligenes sp., Achromobacter sp., porcine pancreas,
and Chromobacterium viscosum. All the lipases were
employed in the immobilized form.
9. This solvent was employed in the lipase-catalyzed highly
enantioselective acylation of 2-aryloxy-1-propanols:
Miyazawa, T.; Yukawa, T.; Koshiba, T.; Sakamoto, H.;
Ueji, S.; Yanagihara, R.; Yamada, T. Tetrahedron:
Asymmetry 2001, 12, 1595.
10. The product distribution after 3 days was as follows: 6b,
24.0%; 7b, 20.0%; 8b, 7.3%.
11. For the product distribution after 3 days, see Table 3.
12. This lipase was found to be highly regioselective as well as
active in the deacylation of resorcinol and hydroquinone
derivatives: see Ref. 4.
of significant importance from a synthetic standpoint,
because either regioisomer of monoacylated derivatives
of dihydroxybenzenes can easily be obtained by choos-
ing either acylation or deacylation mediated by the sin-
gle biocatalyst which is easily available.
References and notes
1. For reviews, see: (a) Faber, K. Biotransformations in
Organic Chemistry, 5th ed.; Springer: Berlin, 2004, pp 94–
123, 344–367; (b) Gais, H. J.; Theil, F. In Enzyme
Catalysis in Organic Synthesis; Drauz, K., Waldmann,
H., Eds., 2nd ed.; Wiley-VHC: Weinheim, 2002; pp 335–
578.
2. Parmar, V. S.; Kumar, A.; Bisht, K. S.; Mukherjee, S.;
Prasad, A. K.; Sharma, S. K.; Wengel, J.; Olsen, C. E.
Tetrahedron 1997, 53, 2163, and references cited therein.
3. Natoli, M.; Nicolosi, G.; Piattelli, M. J. Org. Chem. 1992,
57, 5776.
4. We also have recently reported the highly regioselective
deacylation of resorcinols and hydroquinones acylated at
both phenolic hydroxyls mediated by Candida antarctica
lipase B: Miyazawa, T.; Hamada, M.; Morimoto, R.;
Murashima, T.; Yamada, T. Tetrahedron Lett. 2007, 48,
8334.
5. Nicolosi, G.; Piattelli, M.; Sanfilippo, C. Tetrahedron
1993, 49, 3143.
6. They also reported the formation of two derivatives
protected only at the A ring, that is, 5-O-acetylcatechin
and 7-O-acetylcatechin, in the reaction of (+)-catechin
[(2R,3S)-2-(3,4-dihydroxyphenyl)-3,4-dihydro-2H-1-benz-
opyran-3,5,7-triol] with vinyl acetate in the presence of
supported B. cepacia lipase in acetonitrile: Lambusta, D.;
Nicolosi, G.; Patti, A.; Piattelli, M. Synthesis 1993,
1155.
7. The authentic samples of isomeric monoesters of each
dihydroxybenzene were prepared by enzymatic methods.
For example, both the monopropanoates of t-butylhydro-
quinone (1e) were prepared as follows: 4-O-Propanoyl-t-
butylhdyroquinone (2e) was prepared as the main product
of the lipase-catalyzed direct acylation of 1e (0.5 mmol) as
described in Ref. 15 and purified by preparative TLC on
silica gel using hexane–ethyl acetate (2:1, v/v) as a
developing solvent. On the other hand, the isomer of 2e,
that is, 1-O-propanoyl-t-butylhydroquinone (3e), was
prepared as the main product of the lipase-catalyzed
deacylation of 1,4-di-O-propanoyl-t-butylhydroquinone
(4e) (1.5 mmol) with 2-propanol (4.5 mmol) in diisopropyl
ether at 45 °C and purified in the same manner as above.
The structure of these isomeric monoesters was unambigu-
ously determined by 1H NMR (500 MHz), 13C NMR, and
2D NMR (HMQC, HMBC). Thus, for example, cross-
peaks of the proton of 1-OH with the carbons at C-1, C-2,
13. Enzymatic Reactions in Organic Media; Koskinen, A. M.
P., Klibanov, A. M., Eds.; Blackie A&P: Glasgow, 1996.
14. Boehringer Mannheim Chirazyme L-2, which had a
specific activity of 3.2 U/mg lyophilized powder with
tributyrin at 25 °C.
15. Typical experimental procedure: A solution of a dihydric
phenol (0.1 mmol) and vinyl propanoate (0.3 mmol) in
anhydrous diisopropyl ether (240 ll) was stirred with an
immobilized lipase preparation (40 mg) in a thermostated
incubator. After a certain period of time, the reaction
mixture was filtered through a glass filter and evaporated
to dryness under reduced pressure. The residual oil was
dissolved in DMSO-d6 and subjected to 1H NMR
(500 MHz) analysis for the quantification of the reaction
products. The proton signals in the aromatic region were
mainly employed for the purpose. The whole content of
the reaction mixture was used up for one analysis, and
several discrete reaction mixtures were used at different
reaction times.
16. Under the same reaction conditions 4-t-octylresorcinol
(5c) afforded 46.4% of 6c, 0% of 7c, and 53.6% of 8c.