First asymmetric Diels-Alder reactions of furan and chiral acrylates. Usefulness of acid heterogeneous catalysts

Article abstract of DOI:10.1021/jo961513k

TiCl₄ and ZnCl₂ supported on silica gel catalyze the reactions of furan with methyl, (1R,2S,5R)-menthyl, and (1R,2S,5R)-8-phenylmenthyl acrylates. The best results are obtained when reactions are carried out without a solvent. The endo/exo and diastereofacial selectivities depend on the nature of the catalyst and on the reaction conditions. With the (1R,2S,5R)-menthyl acrylate 44% de in endo cycloadducts and 20% de in exo cycloadducts are the best asymmetric inductions. With the (1R,2S,5R)-8-phenylmenthyl acrylate 68% de in endo and 70% de in exo cycloadducts are obtained.

Full text of DOI:10.1021/jo961513k

J . Org. Chem. 1996, 61, 9479-9482
9479
F ir st Asym m etr ic Diels-Ald er Rea ction s of F u r a n a n d Ch ir a l
Acr yla tes. Usefu ln ess of Acid Heter ogen eou s Ca ta lysts
J ose´ M. Fraile, J ose´ I. Garc´ıa, David Gracia, J ose´ A. Mayoral,* and Elisabet Pires
Departamento de Quı´mica Orga´nica, Instituto de Ciencia de Materiales de Arago´n,
C.S.I.C.-Universidad de Zaragoza, Facultad de Ciencias, E-50009 Zaragoza, Spain
Received August 5, 1996X
TiCl4 and ZnCl2 supported on silica gel catalyze the reactions of furan with methyl, (1R,2S,5R)-
menthyl, and (1R,2S,5R)-8-phenylmenthyl acrylates. The best results are obtained when reactions
are carried out without a solvent. The endo/ exo and diastereofacial selectivities depend on the
nature of the catalyst and on the reaction conditions. With the (1R,2S,5R)-menthyl acrylate 44%
de in endo cycloadducts and 20% de in exo cycloadducts are the best asymmetric inductions. With
the (1R,2S,5R)-8-phenylmenthyl acrylate 68% de in endo and 70% de in exo cycloadducts are
obtained.
In tr od u ction
byproducts and the instability of the cycloadducts makes
the purification difficult. Several methods have been
developed to carry out reactions of furan with nonchiral
acrylates, such as using very long reaction times,9 high
pressure,10,11 ZnI2 at 40 °C in sealed tubes,12 cupric
fluoroborate in the presence of hydroquinone and long
reaction times,13 and Y zeolites.14 However, 79% is the
maximum yield obtained with methyl acrylate at room
temperature and 20 000 atm.10 The situation with chiral
acrylates is even worse given that the Diels-Alder
reactivity of acrylates with furan decreases as the size
of the alcohol increases.11 Finally, the induction of
asymmetry with chiral acrylates is based on the kinetic
control of the Diels-Alder reaction and the easy revers-
ibility observed when furan is used as a diene may have
a nonpredictable effect on the diastereofacial selectivity.
To sum up, the development of a method to carry out
reactions of furan with chiral acrylates with good yields
and selectivities is a very interesting yet difficult task.
2-Substituted 7-oxabicyclo[2.2.1]hept-5-ene derivatives
are powerful intermediates in the preparation of carbo-
hydrates and other biologically active compounds.1-4
These products can be obtained, via Diels-Alder reac-
tions, from furan, an inexpensive compound obtained
from agricultural leftovers.5 Given their relationship
with biologically active compounds, there is a noticeable
interest in the asymmetric preparation of these bicyclic
compounds and excellent results have been described6
using chiral derivatives of acrylonitrile, such as 1-(cy-
anovinyl)-(1′S)-camphanate, as a chiral dienophile. More
recently chiral 2-siloxyfurans have been used as chiral
dienes, with Eu(hfc)3 as a chiral catalyst.7 However,
there is no method for obtaining 7-oxabicyclo[2.2.1]hept-
5-ene-2-carboxylates in an asymmetric way. This fact is
surprising because the racemic compounds have gained
interest as synthetic intermediates3 and because the use
of chiral acrylates is the most widely used methodology
to carry out asymmetric Diels-Alder reactions.8
The lack of references about reactions of chiral acry-
lates with furan is undoubtedly due to the difficulties
related to the use of furan as a diene. Furan is an acid-
sensitive reagent and cannot be used in the presence of
the most common Lewis acids. Heating is often an
ineffective alternative because the cycloadducts are
thermally unstable and easily revert to the reagents or
are cleaved into 2-substituted furans. Furthermore the
use of an excess of diene leads to the formation of
We have shown that several Lewis acids supported on
silica gel are useful catalysts in Diels-Alder reactions
of nonchiral15,16 and chiral16 R,â-unsaturated esters.
Furthermore, some of these supported Lewis acids im-
prove the selectivity, with regard to side reactions, when
used in the absence of a solvent. For instance, they
promote Diels-Alder reactions of nonchiral17 and chiral18
(E)-2-phenyl-5(4H)-oxazolones without causing E/ Z
isomerization.
(9) Nelson, W. L.; Allen, D. R. J . Heterocycl. Chem. 1972, 9, 561.
(10) Dauben, W. G.; Krabbenhoft, H. O. J . Am. Chem. Soc. 1976,
98, 1992.
(11) Kotsuki, H.; Nishizawa, H.; Ochi, M.; Matsuoka, K. Bull. Chem.
Soc. J pn. 1982, 55, 496.
X Abstract published in Advance ACS Abstracts, November 15, 1996.
(1) Vogel, P.; Fattori, D.; Gasparini, F.; Le Drian, C. Synlett 1990,
173.
(2) Schmidt, R.; Beitzke, C.; Forrest, A. W. J . Chem. Soc., Chem.
Commun. 1982, 909.
(3) (a) Campbell, M. M.; Kaye, A. D.; Sainsbury, M. Tetrahedron
Lett. 1983, 24, 4745. (b) Campbell, M. M.; Kaye, A. D.; Sainsbury, M.;
Yavarzadeh, R. Tetrahedron Lett. 1984, 25, 1629.
(4) For more recent examples see for instance: (a) Kerner, P.; Vogel,
P. Helv. Chim. Acta 1995, 78, 301. (b) Kraehenbuehl, K.; Vogel, P.
Tetrahedron Lett. 1995, 36, 8595. (c) Acen˜a, J . L.; Arjona, O.; Iradier,
F.; Plumet, J . Tetrahedron Lett. 1996, 37, 105.
(5) Dean, R. M. Adv. Heterocycl. Chem. 1982, 30, 169.
(6) Vieira, E.; Vogel, P. Helv. Chim. Acta 1983, 66, 1865.
(7) Sha, C.-K.; Shen, C.-Y.; Lee, R.-S.; Lee, S.-R.; Wang, S.-L.
Tetrahedron Lett. 1995, 36, 1283.
(8) (a) Oppolzer, W. Angew. Chem., Int. Ed. Engl. 1984, 23, 876. (b)
Paquette, L. A. In Asymmetric Synthesis; Morrison, J . D., Ed.; Academic
Press: New York, 1984; Vol. 3, p 455. (c) Oppolzer, W. Tetrahedron
1987, 43, 1969. (d) Helchem, G.; Karge, P.; Weetman, J . In Modern
Synthetic Methods; Scheffold, R., Ed.; Springer: Berlin, 1986; Vol. 4,
p 261.
(12) Brion, F. Tetrahedron Lett. 1982, 23, 5299.
(13) Moore, J . A.; Partain, A. M., III. J . Org. Chem. 1983, 48, 1105.
(14) Narayana Murthy, Y. V. S.; Pillai, C. N. Synth. Commun. 1991,
21, 783.
(15) Cativiela, C.; Fraile, J . M.; Garc´ıa, J . I.; Mayoral, J . A.; Pires,
E.; Royo, A. J .; Figueras, F.; de Me´norval, L. C. Tetrahedron 1993, 49,
4073.
(16) (a) Cativiela, C.; Figueras, F.; Garc´ıa, J . I.; Mayoral, J . A.; Pires,
E.; Royo, A. J . Tetrahedron: Asymmetry 1993, 4, 621. (b) Cativiela,
C.; Garc´ıa, J . I.; Mayoral, J . A.; Pires, E.; Royo, A. J . Tetrahedron 1995,
51, 1295. (c) Cativiela, C.; Garc´ıa, J . I.; Mayoral, J . A.; Pires, E.; Royo,
A. J .; Figueras, F. Appl. Catal. A 1995, 131, 159. (d) Garc´ıa, J . I.;
Mayoral, J . A.; Pires, E.; Brown, D. R.; Massam, J . Catal. Lett. 1996,
37, 261. (e) Fraile, J . M.; Garc´ıa, J . I.; Mayoral, J . A.; Pires, E.; Tarnai,
T.; Figueras, F. Appl. Catal. A 1996, 136, 113.
(17) Cativiela, C.; Garc´ıa, J . I.; Mayoral, J . A.; Pires, E.; Brown, R.
Tetrahedron 1995, 51, 9217.
(18) Cativiela, C.; Fraile, J . M.; Garc´ıa, J . I.; Lo´pez, P.; Mayoral, J .
A.; Pires, E. Tetrahedron: Asymmetry 1996, 7, 2391.
S0022-3263(96)01513-7 CCC: $12.00 © 1996 American Chemical Society

9480 J . Org. Chem., Vol. 61, No. 26, 1996
Fraile et al.
Sch em e 1
Ta ble 1. Resu lts Obta in ed in th e Rea ction of F u r a n (1)
Ta ble 2. Resu lts Obta in ed in th e Rea ction of F u r a n (1)
w ith Meth yl Acr yla te (2a ) a fter 24 h a t 25 °C
w ith th e Ch ir a l Acr yla tes 2b a n d 2c a t 25 °C
catalyst
SiO2
1:2a
solvent
CDCl3
% conversiona
endo/exoa,b
% dea
dieno-
time
%
endo/
phile catalyst 1:2 solvent (h) conv exoa endob exob
6:1
6:1
6:1
6:1
6:1
6:1
3:1
6:1
6:1
3:1
0
8
3
51
35
71
52
16
75
22
2b
SiO2
10:1 CH2Cl2
10:1
3
3
3
3
3
3
3
24
24
3
0
0
0
SiO2-Al
SiO2-Ti
CDCl3
CDCl3
60:40
77:23
73:27
50:50
79:21
81:19
86:14
SiO2-Al 10:1 CH2Cl2
10:1
15 50:50
34 66:34
82 62:38
37 57:43
48 55:45
30 76:24
10 68:32
58 50:50
52 55:45
57 50:50
47 50:50
26 31:69
79 33:67
17 60:40
46 50:50
46
32
44
38
43
20
44
32
36
10
26
10
10
10
9
20
8
10
9
19
10
8
SiO2-Ti 10:1 CH2Cl2
10:1
6:1
6:1
3:1
SiO2-Zn
CDCl3
a
b
SiO2-Zn 10:1 CH2Cl2
Determined by 1H-NMR. Assigned on the basis of previously
10:1
6:1
6:1
3:1
3
3
described spectra (ref 9).
24
24
3
24
3
In view of this we have tested these supported Lewis
acids as catalysts in the reaction of furan (1) with methyl,
(1R,2S,5R)-menthyl, and (1R,2S,5R)-8-phenylmenthyl
acrylates (2a , 2b, and 2c, respectively) (Scheme 1).
2c
SiO2-Ti 10:1
10:1
33
68
70
66
SiO2-Zn 10:1
10:1
24
a
b
Determined by 1H- and 13C-NMR. 3n and 3x are the major
Resu lts a n d Discu ssion
cycloadducts.
SiO2-Al and SiO2-Ti were obtained by reaction of
silica gel (Merck, silica gel 60, 63-200 nm) with Et2AlCl
and TiCl4 following the previously described method.15
SiO2-Zn was obtained from ZnCl2 and silica gel (EP11
from Crossfield) as previously described.19 The catalytic
activity of these solids, together with that of silica gel 60
from Merck, was compared in the reaction of furan (1)
with methyl acrylate (2a ). The results obtained (Table
1) show that silica gel does not catalyze this reaction,
whereas the treated silicas do. Supported TiCl4 and
ZnCl2 are more efficient than supported Et2AlCl. The
best chemical yields are obtained when the reactions are
carried out without a solvent and the endo/ exo selectivity
depends on the catalyst, on the proportion of reagents,
and on the presence or absence of solvent. The influence
of the solvent on the endo/ exo selectivity of Diels-Alder
reactions of furan has been also detected in reactions
carried out at high pressure.20
on the basis of their 1H- and 13C-NMR spectra. As in
the adducts obtained from methyl acrylate, the signal
corresponding to H2 appears at a higher chemical shift
in the endo cycloadducts (3n b + 4n b, 3.07 ppm) than in
the exo (3xb + 4xb, 2.42 ppm). The position of the
carbonyl carbons in 13C-NMR agrees with this assign-
ment because in the endo isomers (3n b + 4n b) this
carbon is shielded by the double bond and appears at a
higher field (171.7 ppm) than in the exo isomers (3xb +
4xb, 173.2 ppm). The percentage of conversion and the
different selectivities were detemined by 13C-NMR of the
vinyl carbons [2b, 129.0 ppm (CR), 130.1 ppm (Câ); 3n b
132.3 ppm (C6); 4n b, 132.6 ppm (C6); 4xb, 134.7 ppm (C6);
3xb, 134.8 ppm (C6)].
The cycloadducts cannot be purified by chromatogra-
phy on silica gel because they decompose. This is
probably due to the acidity of the silica gel, located in
the free hydroxyl groups of the surface. If this interpre-
tation is correct, a possible solution to this problem would
be to eliminate most of this acidity through an “end-
capping” procedure. This was effectively carried out by
treatment of the silica gel with hexamethyldisilazane
(HMDS). With this in-house stationary phase, we were
able to successfully separate the endo (3n b + 4n b) and
The same solids were tested as catalysts in the reaction
between furan (1) and (1R,2S,5R)-menthyl acrylate (2b)
(Table 2). The endo and exo cycloadducts were assigned
(19) (a) Rhodes, C. N.; Brown, D. R. J . Chem. Soc., Faraday Trans.
1992, 88, 2269. (b) Rhodes, C. N.; Brown, D. R. J . Chem. Soc., Faraday
Trans. 1993, 89, 1387.
(20) J enner, G. Tetrahedron Lett. 1994, 35, 1189.

Asymmetric Diels-Alder Reactions
J . Org. Chem., Vol. 61, No. 26, 1996 9481
carbons [2c, 129.8 ppm (CR+Câ); 3n c 133.1 ppm (C6); 4n c,
1
132.1 ppm (C6); 3xc + 4xc, 134.5 ppm (C6)] and by H-
NMR of the vinyl protons (2c, 5.75, 5.95, 6.08 ppm
(3Hvinyl); 3n c + 4n c, 6.35 ppm (H6); 4xc 6.30 ppm (H6);
3xc, 6.25 ppm (H6); 3n c + 3xc, 6.20 ppm (H5); 4n c, 6.01
ppm (H5); 4xc, 6.12 ppm (H5)].
Endo (3n c + 4n c) and exo (3xc + 4xc) cycloadducts
were separated by column chromatography on “end-
capped” silica gel using hexane/diethyl ether (100:1) as
an eluent. In this way the major exo cycloadduct (3xc)
was obtained in pure form.
F igu r e 1. Model to account for the direction of the asymmetric
induction experimentally observed.
exo (3xb + 4xb) cycloadducts by column chromatography
using hexane/diethyl ether (100:1) as an eluent.
A mixture of bromo lactones (5 + 6) was obtained, as
described above, from the reaction carried out with SiO2-
In order to determine the absolute configuration of the
major endo cycloadduct, the mixture coming from a
reaction (SiO2-Ti, no solvent, 3 h, 10-fold excess of furan)
was treated with bromine and aqueous sodium hydrogen
carbonate. The mixture of bromo lactones (5 + 6) was
also purified by column chromatography on “end-capped”
silica gel, using hexane/diethyl ether, initially with 100:1
and then with 1:1, as an eluent, followed by recrystalli-
zation from ethyl acetate. The polarimetric analysis of
25
Zn. The mixture obtained gave [R]D ) -70.3° (c 0.99
in CHCl3), indicating that 3n c is the major endo cycload-
duct. This value correponds to 76% ee, so the purification
process increases the amount of the major enantiomer.
This result can be accounted for by the classical model
used in Lewis acid-catalyzed reactions of this dienophile8
(Figure 1), and 3xc was assigned as the major exo
cycloadduct on the basis of this model.
25
The results obtained (Table 2) show that TiCl4 sup-
ported on silica gel (SiO2-Ti) is more efficient than SiO2-
Zn. In contrast to that observed in the reaction of
(1R,2S,5R)-menthyl acrylate (2b), the use of SiO2-Ti
leads to a preference of the exo cycloadducts and further-
more the diastereofacial selectivity in exo cycloadducts
is 70% de. The reaction carried out with SiO2-Zn
showed little or no endo/ exo selectivity, but the diaste-
reofacial selectivity was about the same (67% de) in endo
and exo cycloadducts.
this product gives an [R]D ) -50° (c 0.99 in CHCl3). A
comparison of this value with that previously described21
indicates that 5 is the major bromolactone with 54% ee.
It is important to note that in the purification process
an increase of the ee with regard to that determined in
the crude of the reaction is observed, so that the ee
determined after this purification step cannot be taken
in general as a measure of the asymmetric induction
produced in the reaction.
The formation of 3n b as a major cycloadduct can be
explained by the model used in Lewis acid-catalyzed
reactions of the same dienophile 2b with cyclopentadi-
ene.8 In this model the dienophile displays an anti enoate
conformation where the isopropyl group of the chiral
auxiliary shields the Re face of the dienophile and the
attack of the diene preferentially takes place on the Si
face (Figure 1). On the basis of this model 3xb was
assigned as the major exo cycloadduct.
Con clu sion
It can be concluded that ZnCl2 and TiCl4 supported on
silica gel are fairly efficient catalysts for the reactions of
chiral acrylates with furan. The second catalyst leads
to the best percent conversion, but the endo/ exo and
diastereofacial selectivities depend on the nature of the
catalyst and the chiral auxiliary. The best results are
obtained in reactions carried out in the absence of a
solvent. The purification of the products allows the major
exo cycloadduct (3xc) in pure form and the bromo lactone
with 76% ee to be obtained, and furthermore this enan-
tiomeric excess can be further improved by recrystalli-
zation. The synthetic methodology described here is
capable of being used in the preparation of furan deriva-
tives in enantiomerically pure form, which are of interest
as intermediates in the synthesis of biologically active
compounds. These studies are in progress and will be
reported in due course.
Table 2 summarizes the results obtained from the
reaction of furan (1) and (1R,2S,5R)-menthyl acrylate
(2b). Silica gel and silica gel treated with Et2AlCl are
not good catalysts for this reaction. The best results are
again obtained when the reactions are carried out in the
absence of a solvent. The use of long reaction times does
not noticeably improve the conversion but the selectivity
tends to decrease. Therefore the best results are obtained
by using a large excess of diene with short reaction times.
Under these conditions, the silica modified by treatment
with TiCl4 provides the best results. It is important to
note that the asymmetric induction obtained in the endo
cycloadducts is comparable to that described for the
reaction of this dienophile with cyclopentadiene.16a
The best catalysts, namely SiO2-Ti and SiO2-Zn, were
tested in the reaction between (1R,2S,5R)-8-phenylmen-
thyl acrylate (2c) and furan (1) in the absence of a solvent
(Table 2). The endo and exo cycloadducts were again
assigned by the higher chemical shift of H2 in the endo
(4n c, 2.27 ppm; 3n c, 2.50 ppm) than in the exo (3xc +
4xc, 1.55-1.70 ppm) and by the lower chemical shift of
the carbonyl carbon in the endo (4n c, 171.2 ppm; 3n c,
171.6 ppm) than in the exo cycloadducts (3xc + 4xc,
173.2 ppm). The percentage of conversion and the
selectivities were determined by 13C-NMR of the vinyl
Exp er im en ta l Section
(1R,2S,5R)-Menthyl and (1R,2S,5R)-8-phenylmenthyl acry-
lates were prepared according to a procedure described in the
literature.22 Silica-supported AlEt2Cl and TiCl4 were prepared
from Merck silica gel 60 as previously described.15 ZnCl2 was
supported on EP11 silica from Crossfield following a previously
described procedure.19 The silicas contained 1.4 mmol of Al
g-1, 1.2 mmol of Ti g-1, and 1.5 mmol of Zn g-1, respectively.
Diels-Ald er Rea ction s in Solu tion . Under argon, 1
mmol of the dienophile (2a or 2b) and the corresponding
amount of freshly distilled furan (6 mmol for 2a and 10 mmol
for 2b) were added to a suspension of the catalyst (1 g) in 6
mL of the solvent (CDCl3 for 2a and CH2Cl2 for 2b). The
(21) Ogawa, S.; Iwasawa, Y.; Nose, T.; Suami, T.; Ohba, S.; Ito, M.;
Saito, Y. J . Chem. Soc., Perkin Trans. 1 1985, 903.
(22) Oppolzer, W.; Kurth, M.; Reichlin, D.; Chapuis, C.; Mohnhaupt,
M.; Moffat, F. Helv. Chim. Acta 1981, 64, 2802.

9482 J . Org. Chem., Vol. 61, No. 26, 1996
Fraile et al.
reactions were shaken for the time indicated (Tables 1 and
(1R,2S,5R)-8-P h en ylm en th yl (1S,2R,4S)-7-Oxa bicyclo-
[2.2.1]h ep t-5-en e-2-exo-ca r boxyla te (3xc). This product
was separated from the reaction mixture as described above.
1H-NMR (CDCl3, 300 MHz) δ: 7.3-7.2 (m, 3H), 7.2-7.1 (m,
2H), 6.25 (dd, J ) 6.1, 1.6 Hz, 1H), 6.20 (dd, J ) 6.1, 1.6 Hz,
1H), 4.95 (m, 1H), 4.90 (m, 1H), 4.85 (m, 1H), 2.01 (m, 1H),
1.82 (m, 1H), 1.70-1.55 (m, 4H), 1.5-1.4 (m, 1H), 1.23 (s, 3H),
1.20 (m, 1H), 1.15 (s, 3H), 1.15-0.9 (m, 3H), 0.85 (d, 3H). 13C-
NMR (CDCl3, 75 MHz) δ: 173.2, 151.8, 136.6, 134.5, 127.8,
125.4, 124.9, 80.0, 77.1, 74.4, 50.0, 42.2, 41.6, 39.6, 34.5, 31.2,
29.4, 28.4, 26.4, 24.5, 21.7. Anal. Calcd for C23H30O3: C, 77.93;
H, 8.65. Found C, 77.78; H, 8.79.
(1R,2S,5R)-8-P h en ylm en th yl (1R,2S,4R)-7-Oxa bicyclo-
[2.2.1]h ep t -5-en e-2-exo-ca r b oxyla t e (4xc). The data of
this product were obtained from a mixture of 3xc and 4xc.
1H-NMR (CDCl3, 300 MHz) δ: 7.3-7.2 (m, 3H), 7.2-7.1 (m,
2H), 6.30 (dd, J ) 5.9, 1.7 Hz, 1H), 6.12 (dd, J ) 5.9, 1.7 Hz,
1H), 4.95 (m, 1H), 4.85 (m, 1H), 4.58 (m, 1H), 2.01 (m, 1H),
1.82 (m, 1H), 1.70-1.55 (m, 4H), 1.5-1.4 (m, 1H), 1.23 (s, 3H),
1.20 (m, 1H), 1.15 (s, 3H), 1.15-0.9 (m, 3H), 0.85 (d, 3H). 13C-
NMR (CDCl3, 75 MHz) δ: 173.2, 151.8, 136.6, 134.5, 127.8,
125.4, 124.9, 80.0, 77.1, 74.4, 50.0, 42.2, 41.6, 39.6, 34.5, 31.2,
29.4, 28.4, 26.4, 24.5, 21.7. Anal. Calcd for C23H30O3: C, 77.93;
H, 8.65. Found C, 77.89; H, 8.74.
(1R,2R,3R,6R,7S)- a n d (1S,2S,3S,6S,7R)-2-Br om o-4,8-
d ioxa tr icyclo[4.2.1.03,7]n on a n -5-on e (5 + 6). The crude
obtained from a reaction of 2b or 2c was stirred with a solution
of NaHCO3 (1.6 mmol) and Br2 (2.4 mmol) in water (3 mL) for
1 h at room temperature. After this time the aqueous solution
was extracted with ethyl acetate, the organic phase was
washed with a saturated solution of Na2S2O3 and dried over
anhydrous Na2SO4, and the solvent was evaporated under
reduced pressure. The oil obtained was purified by column
chromatography using “end-capped” silica gel as a stationary
phase and hexane/diethyl ether (initially with 100:1 and then
with 1:1) as an eluent. The different fractions obtained were
monitored by gas chromatography (FID detector on a Hewlett-
Packard 5890II, cross-linked methyl silicone column 25 m ×
0.2 mm × 0.33 µm, helium as carrier gas, 20 psi, injector
temperature 230 °C, detector temperature 250 °C, oven
temperature program 100 °C (2 min)s25 °C/mins220 °C (5
min), retention times: (-)-menthol 5.4 min, bromolactone 7.6
min, (-)-8-phenylmenthol 10.5 min). In this way the bromo
lactone was separated and the chiral auxiliaries were recov-
ered. The bromo lactone was further purified by recrystalli-
zation from ethyl acetate. 1H-NMR (CDCl3, 300 MHz) δ: 5.41
(t, J ) 4.8 Hz), 4.95 (d, J ) 4.8 Hz), 4.75 (d, J ) 5.1 Hz), 3.92
(s), 2.75 (m), 2.30 (m), 2.09 (dd, J ) 13.5, 1.8 Hz). 13C-NMR
(CDCl3, 75 MHz) δ: 175.8, 86.1, 83.0, 81.5, 50.3, 38.2, 35.6.
Anal. Calcd for C7H7BrO3: C, 38.39; H, 3.22. Found C, 38.18;
H, 3.10.
2). For the reactions of 2a the solution was extracted from
1
the reaction flask with a syringe and directly analyzed by H-
NMR. For the reactions of 2b the catalyst was separated by
filtration and washed with CH2Cl2. The solvent was elimi-
nated under reduced pressure and the crude reactions were
1
analyzed by H- and 13C-NMR in CDCl3. The conversion and
selectivities were determined by integration of the signals
given in the text.
Diels-Ald er Rea ction s in th e Absen ce of a Solven t.
Under argon, 1 mmol of the dienophile (2a , 2b, or 2c) and the
corresponding amount of freshly distilled furan (Tables 1 and
2) were added to 1 g of the catalyst. The reactions were stirred
for the time indicated (Tables 1 and 2). For the reactions of
2a , 6 mL of CDCl3 were added, and the solution was extracted
with a syringe and analyzed by 1H-NMR. For the reactions
of 2b and 2c, 6 mL of CH2Cl2 were added, and the catalyst
was separated by filtration and thoroughly washed with CH2-
Cl2. The solvent was eliminated under reduced pressure and
the crude reactions were analyzed by 1H- and 13C-NMR in
CDCl3. The conversion and selectivities were determined by
integration of the signals given in the text.
En d -Ca p p in g of th e Silica Gel. From a suspension of
silica gel (100 g) in toluene (400 mL), a mixture of toluene and
water (100 mL) was distilled. Then 50 mL of hexamethyl-
disilazane (HMDS) were added, and the mixture was heated
under reflux for 45 min. After this time, the solid was
separated by filtration and thoroughly washed with diethyl
ether, ethanol, and diethyl ether and then dried under vacuum.
This solid was used in the chromatographic purification of the
cycloadducts.
(1R,2S,5R)-Men th yl (1R,2R,4R)- a n d (1S,2S,4S)-7-Oxa -
bicyclo[2.2.1]h ept-5-en e-2-en do-car boxylates (3n b + 4n b).
These products were separated from the reaction mixture by
column chromatography using the “end-capped” silica gel and
hexane/diethyl ether (100:1) as an eluent. 1H-NMR (CDCl3,
300 MHz) δ: 6.42 (dd, J ) 5.7, 1.8 Hz), 6.20 (dd, J ) 5.7, 1.8
Hz), 6.15 (dd, J ) 5.7, 1.5 Hz), 5.15 (m), 5.01 (m), 4.60 (m),
3.07 (m), 2.12-2.02 (m), 1.95-1.85 (m), 1.85-1.75 (m), 1.7-
1.6 (m), 1.55 (m), 1.5-1.3 (m), 1.1-0.9 (m), 0.85 (m), 0.7 (m).
13C-NMR (CDCl3, 75 MHz) δ: 171.7, 137.2, 137.0, 132.6, 132.3,
79.0, 78.9, 74.6, 46.9, 43.1, 40.8, 34.2, 31.3, 28.2, 26.1, 23.1,
22.0, 20.8, 15.9. Anal. Calcd for C17H26O3: C, 73.35; H, 9.32.
Found C, 73.29; H, 9.20.
(1R,2S,5R)-Men th yl (1S,2R,4S)- a n d (1R,2S,4R)-7-Oxa -
bicyclo[2.2.1]h ep t-5-en e-2-exo-ca r boxyla tes (3xb + 4xb).
These products were separated from the reaction mixture as
described above. 1H-NMR (CDCl3, 300 MHz) δ: 6.38 (m), 5.15
(m), 5.05 (m), 4.71 (m), 2.42 (m), 2.2-2.1 (m), 2.05-1.95 (m),
1.9-1.8 (m), 1.7-1.6 (m), 1.5-1.3 (m), 1.1-0.9 (m), 0.85 (m),
0.7 (m). 13C-NMR (CDCl3, 75 MHz) δ: 173.2, 137.0, 137.0,
134.8, 134.7, 81.1, 78.0, 74.6, 47.0, 43.1, 40.8, 34.2, 31.3, 29.0,
26.2, 23.3, 22.0, 20.8, 16.1. Anal. Calcd for C17H26O3: C, 73.35;
H, 9.32. Found C, 73.18; H, 9.60.
Ack n ow led gm en t. This work was made possible by
the generous financial support of the Comisio´n Inter-
ministerial de Ciencia y Tecnolog´ıa (project MAT96-
1053-C02-01). One of us (E.P.) thanks the Diputacio´n
General de Arago´n for a grant.
(1R,2S,5R)-8-P h en ylm en th yl (1R,2R,4R)- an d (1S,2S,4S)-
7-Oxa bicyclo[2.2.1]h ep t-5-en e-2-en d o-ca r boxyla tes (3n c
+ 4n c). These products were separated from the reaction
mixture by the same chromatographic method described above
for the cycloadducts obtained from 2b. 1H-NMR (CDCl3, 300
MHz) δ: 7.3-7.2 (m), 7.2-7.1 (m), 6.35 (dd, J ) 5.7, 1.8 Hz),
6.20 (dd, J ) 6.1, 1.6 Hz), 6.01 (dd, J ) 5.7, 1.6 Hz), 4.92 (m),
4.85 (m), 4.72 (m), 4.50 (m), 2.50 (m), 2.27 (m), 2.05-1.95 (m),
1.9-1.8 (m), 1.7-1.55 (m), 1.45-1.35 (m), 1.30 (s), 1.21 (m),
1.19 (s), 1.1-0.9 (m), 0.8 (d). 13C-NMR (CDCl3, 75 MHz) δ:
171.6, 171.2, 151.5, 136.4, 133.1, 132.1, 127.9, 125.3, 125.1,
78.7, 78.6, 74.5, 50.0, 43.0, 41.6, 39.6, 34.4, 31.2, 29.2, 26.5,
24.7, 23.3, 21.7. Anal. Calcd for C23H30O3: C, 77.93; H, 8.65.
Found C, 78.10; H, 8.66.
Su p p or tin g In for m a tion Ava ila ble: A full listing of 1H-
and 13C-NMR data of compounds 3n b + 4n b, 3xb + 4xb, 3n c
+ 4n c, 3xc, 4xc, and 5 + 6, complete with peak assignments
(2 pages). This material is contained in libraries on microfiche,
immediately follows this article in the microfilm version of the
journal, and can be ordered from the ACS; see any current
masthead page for ordering information.
J O961513K