Synthesis of optically active 2,3-dihydrobenzofuran derivatives through a combination strategy of iron(III)-catalyzed reaction and enzymatic reaction

Article abstract of DOI:10.1016/S0040-4039(03)00843-8

Synthesis of optically active 2,3-dihydrofuran derivatives has been accomplished through a combination strategy of ferric ion-catalyzed cycloaddition of styrene derivatives with quinones and following lipase-catalyzed enantioselective acylation.

Full text of DOI:10.1016/S0040-4039(03)00843-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 4081–4084
Synthesis of optically active 2,3-dihydrobenzofuran derivatives
through a combination strategy of iron(III)-catalyzed reaction
and enzymatic reaction
Toshiyuki Itoh,* Kimio Kawai, Shuichi Hayase and Hiroyuki Ohara
Department of Materials Science, Faculty of Engineering, Tottori University, Tottori 680-8552, Japan
Received 7 March 2003; revised 27 March 2003; accepted 28 March 2003
Abstract—Synthesis of optically active 2,3-dihydrofuran derivatives has been accomplished through a combination strategy of
ferric ion-catalyzed cycloaddition of styrene derivatives with quinones and following lipase-catalyzed enantioselective acylation.
© 2003 Elsevier Science Ltd. All rights reserved.
Iron is recognized as an economical and pollution free
metal source.1 Recently we reported the synthesis of
2,3-dihydrobenzofuran by the reaction of styrene
derivatives with 1,4-benzoquinone using alumina-sup-
ported iron(III) perchlorate or iron(II) tetra-
fluoroborate as catalysts,2 and we found that the
reaction was greatly accelerated in an ionic liquid sol-
vent system.3 Since 2,3-dihydrobenzofuran molecular
flame is common in natural compounds,4 it was neces-
sary to establish the synthetic route to access optically
active form of 2,3-dihydrobenzofuran derivatives.
Engler and co-workers reported excellent results in the
enantioselective synthesis of 2,3-dihydrobenzofuran
derivatives using chiral Lewis acid-catalyzed reaction;5
however, a large excess amount of a chiral Titanium-
(IV) complex was required for the reaction.5,6 From the
standpoint of Green Chemistry, we should develop
a greener reaction process through catalytic reaction
systems. So, we wanted to synthesize optically active
2,3-dihydrobenzofuran derivatives through a catalytic
reaction process. Here we report the successful results
of the synthesis of optically pure 2,3-dihydrobenzo-
furan derivatives using a combination strategy of iron
salt-catalyzed reaction and enzymatic reaction as illus-
trated in Scheme 1.
Since iron(III) salt-catalyzed cycloaddition of styrene
derivatives with 1,4-benzoquinone was strongly depen-
dent on the nature of these derivatives,2 we first tested
the cycloaddition reaction using 3-(4-methoxy)phenyl-
2-propenol (2a) as substrate in the presence of 3 mol%
of alumina-supported iron(III) perchlorate in acetoni-
trile (Eq. (1)). But no desired compound was obtained
and starting 2a was recovered after 24 h reaction as
shown in Table 1 (entry 1).
Iron(II) tetrafluoroborate worked very well for the
reaction of trans-anethol with 1,4-benzoquinone in the
[bmim]PF6 solvent system,7 but no desired product was
obtained for the reaction of 2a in this catalytic system
either (entry 2). t-Butyldimethylsilyl ether 2c and allylic
ether 2d also gave no desired product (entries 3 and 4)
and it was finally found that the hydroxyl group must
be protected as acetate; 2-(4-methoxy)phenyl-3-
acetoxymethyl-2,3-dihydro-5-benzofuranol (1b) was
obtained with high trans selectivity (11:1) from 3-(4-
methoxy)phenyl-2-propenyl acetate (2b) in 84% yield
(entry 4). The reaction of 2b using iron(II) tetra-
fluoroborate as catalyst in [bmim]PF6 proceeded, but
Scheme 1.
* Corresponding author. E-mail: titoh@chem.tottori-u.ac.jp
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)00843-8

4082
T. Itoh et al. / Tetrahedron Letters 44 (2003) 4081–4084
Table 1. Synthesis of 2-aryl-3-hydroxymethyl-2,3-dihydrobenzofuran derivatives using iron salt-catalyzed reaction
(1)
Entry
R
Catalyst
Solvent
Time
Yield (%)a
trans/cisb
1
2
3
4
5
6
H
H
Fe(ClO4)3/Al2O3
Fe(BF4)2/6H2O
Fe(ClO4)3/Al2O3
Fe(ClO4)3/Al2O3
Fe(ClO4)3/Al2O3
Fe(BF4)2/6H2O
CH3CN
[bmim]PF6
CH3CN
CH3CN
CH3CN
24 h
1 h
48 h
24 h
2 h
0c
0c
–
–
–
–
11:1
20:1
SiMe2But
CH2CHꢀCH2
Ac
0c
0d
84
23
Ac
[bmim]PF6
20 min
a Isolated yield.
b Determined by capillary GC-analysis.
c No reaction took place and starting compound 2 was recovered.
d Unidentified polymeric products were produced but neither desired product 1 nor starting compound 2 was recovered.
matic acetylation of ( )-1a proceeded smoothly in
diisopropyl ether at 35°C (Eq. (2)); however, the best
enantioselectivity (E value10) was only 10 when
Novozyme 435 was used as catalyst.11,12 This was
insufficient from a practical aspect. We anticipated that
protection of the phenolic hydroxyl group on the C-
ring of ( )-1a might affect the enantioselectivity of the
enzymatic reaction, because the acidic phenol hydroxyl
group is assumed to bind with a peptide residue of the
enzyme through a hydrogen bond formation; this may
cause an incorrect orientation of the large aromatic ring
moiety in the active site. Three types of compounds,
the result was insufficient (entry 5). Therefore, alumina-
supported iron(III) perchlorate in acetonitrile was
appropriate for this type of substrate, while iron(II)
catalyst in an ionic solvent system gave better results
when trans-anethol was used as substrate.2
Next we attempted to prepare optically active 2,3-dihy-
drobenzofuran by lipase-catalyzed enantioselective
transesterification.9 Acetate ( )-1b was treated with
potassium carbonate in methanol to release alcohol
( )-1a, which was subjected to the lipase-catalyzed
acylation using vinyl acetate as an acyl donor. Enzy-
Table 2. Synthesis of optically active 2-aryl-3-hydroxymethyl-2,3-dihydrofuran derivatives
(2)
Entry
R1
Enzymea
Time (h)
Yield of 4b (%
[h]D of 4 (c ca. 1, CHCl3)
Yield of 1b (% ee)c Conv.d E valuee
ee)c
1
2
3
4
5
6
7
8
9
H
H
H
Me
Me
Me
Me
Novozyme
PS
SL
Novozyme
PS
5
26
48
0.75
4
10
3
1
21 (78)
33 (45)
18 (65)
47 (91)
50 (2)
−12
−6
−3
−58
−3
+51
+1
38 (21)
57 (35)
54 (22)
32 (79)
33 (10)
69 (38)
64 (13)
60 (\99)f
50 (69)
0.21
0.44
0.25
0.47
0.85
0.32
0.64
0.61
0.41
10
4
6
49
1
MY
QL
13 (82)
28 (8)
15
1
MOM Novozyme
Bn Novozyme
25 (64)
−28
−36
\100
\400
2
40 (\99)f
a See reference section.
b Isolated yield.
c Enantiomeric excess was determined by HPLC analysis (Daicel ChiralpakOD, hexane/i-PrOH=8:1 or 20:1, 1.0 ml/min).
d conv. (c) was calculated by the following formula: c=eeS/(eeS=eeP).10
e E=ln[1−c(1+eeP)]/ln[1−c(1−eeP)], here c means conv.
f Only one enantiomer was detected by HPLC analysis.

T. Itoh et al. / Tetrahedron Letters 44 (2003) 4081–4084
4083
5-methoxybenzofuran ( )-1e,8 5-methoxymethylbenzo-
furan ( )-1f, and 5-benzyloxybenzofuran ( )-1g were
prepared,13 and subjected to enzymatic transesterifica-
tion (Eq. (2)). As we anticipated, remarkable enhance-
ment in enantioselectivity was accomplished using these
three substrates in the enzymatic transesterification
(Table 2).
Scientific Research on Priority Areas (A) ‘Exploitation
of Multi-Element Cyclic Molecules’ from the Ministry
of Education, Culture, Sports, Science and Technology,
Japan. The authors are grateful to Novo Nordisk
Bioindustry Ltd, Meito Sangyo Ltd and Amano
Enzyme Ltd for providing lipases.
Practical optical resolution was thus realized when
methoxy derivative ( )-1e, methoxymethyl derivative
( )-1f, or benzyloxy derivative ( )-1g was used as sub-
strate. The lipase-catalyzed acylation worked very well
with the preference of the (2S,3S)-enantiomer,17 and
the E values almost reached a sufficient level (entries 4,
8, and 9). Remarkable acceleration was also achieved
when 5-methoxy derivative ( )-1e was subjected to the
Novozyme-catalyzed reaction, and acetate (2S,3S)-4e
was obtained in 47% yield with 91% ee after just 0.75 h
(entry 4), while it took 5 h when ( )-1a was subjected to
the reaction (entry 1). It should be emphasized that
perfect enantioselective reaction was accomplished
when benzyl protected ( )-1g was used as a substrate;
enantiomerically pure (>99% ee) acetate (2S,3S)-4g was
obtained in 40% (80% theoretical yield) yield after 2 h
reaction at rt (entry 9). In the case of lipase PS-cata-
lyzed reaction, no enhancement in enantioselectivity
was recorded by protection of the phenolic hydroxyl
group at the 5 position, although remarkable accelera-
tion in the acylation was achieved (entry 5). Lipase MY
was the second choice of this reaction among tested
enzymes, though this enzyme has an opposite enan-
tiomer preference for this substrate and (2R,3R)-4e was
produced by lipase MY-catalyzed reaction (entry 6).
References
1. For a review, see: Laszlo, P. Acc. Chem. Res. 1986, 19, 121.
2. Ohara, H.; Kiyokane, H.; Itoh, T. Tetrahedron Lett. 2002,
43, 3041.
3. For recent reviews of ionic liquid solvent system, see:
Welton, T. Chem. Rev. 1999, 99, 2071; Ionic Liquids–Indus-
trial Applications to Green Chemistry; Rogers, R. D.;
Seddon, K. R., Eds.; ACS symposium series 818, Oxford
University Press, 2002.
4. For examples, see: (a) Gregson, M.; Gottlieb, O. R.
Phytochemistry 1978, 17, 1395; (b) Donnelly, B. J.; Don-
nelly, D. M. X.; O’Sulllivan, A. M.; Prendergast, J. P.
Tetrahedron 1969, 25, 4409.
5. (a) Engler, T. A.; Letavic, M. A.; Reddy, J. P. J. Am. Chem.
Soc. 1991, 113, 5068; (b) Engler, T. A.; Letavic, M. A.;
Iyengar, R.; LaTessa, K. O.; Reddy, J. P. J. Org. Chem.
1999, 64, 2391.
6. (a) Engler, T. A.; Combrink, K. D.; Letavic, M. A.; Lynch,
K. O., Jr.; Ray, J. E. J. Org. Chem. 1994, 59, 6567; (b)
Engler, T. A.; Combrink, K. D.; Ray, J. E. J. Am. Chem.
Soc. 1988, 110, 7931.
7. Huddleston, J. G.; Willauer, H. D.; Swatloski, R. P.;
Visser, A. E.; Rogers, R. D. Chem. Commun. 1998, 1765.
8. Typical experimental procedure for preparing ( )-1e: To a
solution of 3-(4-methoxy)phenyl-2-propene-1-yl acetate
(1.14 g, 5.5 mmol) and 1,4-benzoquinone (654 mg, 6.05
mmol) in dry-acetonitrile (1.65 ml) was added Fe(ClO3)/
Al2O3 (625 mg, 0.06 equiv.) in one portion, and the mixture
was stirred at rt for 1 h, then filtered through a florisil short
column to give ( )-1b (1.44 g, 84%). This acetate was
dissolved in a mixed solvent of methanol (15 ml) and
acetone (15 ml), then potassium carbonate (0.736 g, 1.1
equiv.) powder was added and the mixture was stirred for
12 h at rt. The resulting mixture was filtered through a
florisil short column and evaporated to dryness. The
resulting residue was dissolved in methanol (24 ml) and
methyl iodide (2.55 g, 18 mmol) and potassium carbonate
(1.38 g, 10 mmol) were added. The mixture was stirred for
5 days at rt, evaporated to dryness and purified by silica
gel flash column chromatography to give ( )-[5-methoxy-2-
(4-methoxyphenyl)-2,3-dihydrobenzofuran-3-yl]methanol
(1e) (1.07 g, 81%). Selected spectra data for ( )-1e: Bp
240°C, 1.6 Torr/Kugelrohr; Rf 0.6 (hexane/ethyl acetate=
1:1); IR (neat) 3364, 2835, 1614, 1514, 1487, 1250, 1176
cm−1; 1H NMR (270 MHz, CDCl3, l, ppm) 1.67 (1H, brs,
OH), 3.44 (1H, dt, J=8.7 Hz, 5.6 Hz), 3.68 (3H, s), 3.71
(3H, s), 3.80 (2H, dd, J=8.6 Hz, 5.6 Hz), 5.43 (1H, d,
J=6.6 Hz), 6.60–6.71 (3H, m), 6.79 (2H, d, J=8.96 Hz),
7.22 (2H, d, J=8.2 Hz); 13C NMR (125 MHz, CDCl3, l)
30.72, 53.51, 55.09, 55.81, 64.02, 86.70, 109.29, 110.66,
113.63, 113.83, 127.04, 127.53, 127.70, 128.21, 133.66,
153.82, 154.03, 159.20 ppm. Anal calcd for C, 71.31; H,
6.34. Found for C, 71.55; H, 6.32.
It was very interesting that enantioselectivity of the
enzymatic reaction was strongly influenced by the sub-
stituent apart from the reaction point in the substrate.
We are now assuming that protection of the 5-hydroxyl
group is responsible for orienting the aromatic group in
the incorrect cavity in the active site.18 We tend to focus
only on the functional groups located near to the
reaction point in designing a suitable substrate molecule
for the enzymatic reaction. However, the present results
suggest there is a possibility of improving enantioselec-
tivity of the enzymatic reaction by proper modification
of the substrate, even if the original reaction was inade-
quate one.
In conclusion, we accomplished the synthesis of opti-
cally active 2,3-dihydrobenzofuran derivatives through
a combination strategy of ferric iron-catalyzed reaction
and enzymatic reaction. It should be noted that the
enantioselectivity of lipase-catalyzed reaction was
remarkably modified by protecting the phenolic
hydroxyl group on the C-ring which is located apart
from the reaction point. Further investigation of the
scope and limitations of this reaction will make it even
more beneficial.
Acknowledgements
This research was supported by a Grant-in-Aid for

4084
T. Itoh et al. / Tetrahedron Letters 44 (2003) 4081–4084
12. Enantiomeric excess was determined as corresponding
acetate by HPLC analysis (Diacel Chiralpak OD, hexane/
i-PrOH=8:1, 1.0 ml/min; Rt 8.6 min ((2S,3S)-4e), Rt 9.7
min ((2R,3R)-4e).
13. Preparation of 1e, 1f, and 1g was accomplished as shown
in Scheme 2.
14. Mackenzie, A. R.; Moody, C. J.; Ress, C. W. Tetra-
hedron 1986, 42, 3259.
15. Rall, G. J. H.; Oberholzer, M. E.; Ferreora, D.; Roux, D.
G. Tetrahedron Lett. 1976, 17, 1033.
16. Schmidhammer, H.; Brossi, A. J. Org. Chem. 1983, 48,
1469.
17. Absolute configuration of benzofuran (−)-4e (60% ee) was
tentatively assigned to be (2S,3S) by indication of the
sign of specific rotation of 5e compared with the reported
specific rotation value of (2S,3S)-2-methyl-3-(4-
methoxy)phenyl-6-methoxy-2,3-dihydro-5-benzofuranol
(6)5b (Eq. (3-1)). Absolute configuration of benzofuran
(+)-1e (87% ee) was assigned to be (2R,3R) by the same
manner (Eq. (3-2)).
(3-1)
Scheme 2.
9. We initially tested lipase-catalyzed hydrolysis of acetate
( )-1b; however, the enantioselectivity was insufficient;
the best enantioselectivity (E value10) was 11 for
Novozyme 435 among tested nine enzymes. In addition,
protection of the phenolic hydroxyl group of ( )-1b was
unsuccessful and formed deacetylated ( )-1a. We there-
fore decided to investigate lipase-catalyzed transesterifica-
tion of ( )-1a.
(3-2)
10. Chen, C.-S.; Fujimoto, Y.; Girdauskas, G.; Sih, C. J. J.
Am. Chem. Soc. 1982, 102, 7294.
11. Lipase PS (Pseudomonas cepacia, Amano), Lipase MY
(Candida rugosa, Meito), Lipase QL (Alcaligenes sp.,
Amano), Lipase SL (Pseudomonas cepacia SL25, Meito),
Novozyme 435 (Candida antarctica), PPL (Porcine pan-
creatic lipase, Sigma), and LIP (Bacillus sp., Toyobo)
were tested. PPL and LIP did not work with this
substrate.
18. Another explanation may be possible: the poor enantiose-
lectivity is due to the intramolecular or intermolecular
acyl transfer reaction and protection of the hydroxyl
group makes impossible such acyl transfer reaction
between 2-hydroxymethyl group and 5-hydroxyl group in
the reaction course.