Diels-Alder reaction of sugar-derived cyclic dienophiles with cyclopentadiene. Factors affecting the reactivity and stereoselectivity

Article abstract of DOI:10.1021/ol005619k

A sugar-derived cyclic dienophile showed a high exo and face selectivity in the Diels-Alder reaction with cyclopentadiene. Structural modifications on the dienophile structure were made, and the outcomes of the cycloaddition reactions were investigated. T

Full text of DOI:10.1021/ol005619k

ORGANIC
LETTERS
2000
Vol. 2, No. 8
1073-1076
Diels−Alder Reaction of Sugar-Derived
Cyclic Dienophiles with
Cyclopentadiene. Factors Affecting the
Reactivity and Stereoselectivity
Silvina C. Pellegrinet and Rolando A. Spanevello*
Instituto de Qu´ımica Orga´nica de S´ıntesis, Facultad de Ciencias Bioqu´ımicas y
Farmace´uticas, UniVersidad Nacional de Rosario-CONICET, Casilla de Correo 991,
Suipacha 531, 2000 Rosario, Argentina
rspaneVe@fbioyf.unr.edu.ar
Received February 3, 2000
ABSTRACT
A sugar-derived cyclic dienophile showed a high exo and face selectivity in the Diels−Alder reaction with cyclopentadiene. Structural modifications
on the dienophile structure were made, and the outcomes of the cycloaddition reactions were investigated. These results and the spectroscopic
data recorded within a series of analogues were used to analyze the factors affecting the reactivity and exo selectivity.
The Diels-Alder reaction has been widely applied to the
syntheses of complex natural products. The reaction itself
has become a valuable tool in the development of synthetic,
mechanistic, and theoretical concepts, and this important role
could be attributed to its remarkable regio- and stereoselec-
tivities. It is generally accepted that the addition of cyclic
dienes to cyclic dienophiles is the case in which the endo
rule is most applicable; nevertheless, there are examples in
the literature of Diels-Alder reactions that violate this rule.1
The use of carbohydrates as a source of chirality has been
thoroughly described in the literature,2,3 and they have also
proved to be useful raw materials for the preparation of either
diene or dienophiles4 as intermediates in the synthesis of
optically pure compounds. Despite a few exceptions,5,6 it has
been established that cyclic sugar-derived dienophiles usually
follow the Alder rule. Figure 1 shows a different possible
Figure 1. Types of cyclic dienophiles.
interpretation for the term cyclic dienophile in which n is a
number that does not allow the system to adopt a different
conformation.
(1) Fotiadu, F.; Archavlis, A.; Buono, G. Tetrahedron Lett. 1990, 31,
4863.
(2) Collins, P.; Ferrier, R. Monosaccharides; John Wiley & Sons, Inc:
New York, 1995.
(3) Bols, M. Carbohydrate Building Blocks; John Wiley & Sons, Inc.:
New York, 1996.
(4) Cycloaddition Reactions in Carbohydrate; Chemistry; Giuliano, R.
M. Ed.; ACS Symposium Series 494; American Chemical Society:
Washington, DC, 1992.
(5) Rehnberg, N.; Sundin, A.; Magnusson, G. J. Org. Chem. 1990, 55,
5477.
(6) Kim, K. S.; Cho, I. H.; Joo, Y. H.; Yoo, I. Y.; Song, J. H.; Ko, J. H.
Tetrahedron Lett. 1992, 33, 4029.
10.1021/ol005619k CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/29/2000

In this context, we came across a cyclic dienophile system
17 that could be classified as unconstrained (Type III). This
dienophile showed an exceptional exo and face selectivity
when it was allowed to react with cyclopentadiene, affording
adduct 2 as a single product (Scheme 1).8 The subsequent
in Scheme 2.13 It is noteworthy to mention that the reaction
of 1 with ethylmagnesium bromide also afforded the primary
Scheme 2
Scheme 1
cleavage of the cycloadduct double bond unveiled a highly
functionalized cyclopentane ring, which was proposed as a
versatile building block for several different natural products.9
More recently, we proved the feasibility of its transformation
into the angularly fused quinane skeleton of the antibiotic
pentalenolactone.10
The high exo selectivity observed with 1 is an interesting
example of an abnormal addition mode. Unfortunately, the
factors that determine the outcome of a cycloaddition process
can hardly ever be established. Even though in previous
studies Rousch11 and Takeda12 attributed the predominant
exo selectivity of some conformationally rigid cyclic s-cis
dienophiles to dipolar effects, in our case we could not
disregard the influence of steric or stereoelectronic effects
in their multiple forms. We hence decided to study the factors
affecting the stereoselectivity of this reaction by making
structural modifications on the original dienophile system
and comparing the outcome of the resulting cycloaddition.
Our first attempt involved the elongation of the exocyclic
hydrocarbon side chain by synthesizing the corresponding
methyl and ethyl ketones (3 and 8). This change should
increase the steric hindrance of the side chain and reveal
whether it has any influence on the exo selectivity. The
syntheses of ketones 3 and 8 from aldehyde 1 are outlined
alcohol 7 as a byproduct. The reduction of the carbonyl group
is known to compete with the addition process when the
carbonyl group is sterically hindered. Furthermore, in the
1H NMR spectrum of ethyl ketone 8, the signal for the
methylene group R to the carbonyl appeared as four quartets.
This coupling pattern can only be explained if the methylene
group has a restricted rotation around the single bond so both
protons are nonequivalent, which confirmed the steric
congestion around the carbonyl group.
The Diels-Alder reaction of the methyl ketone 2 afforded
only 12% yield of a mixture of exo-â/endo-â cycloadducts
(4 and 5)14 in a 2.2:1 ratio, even after 48 h. Under the same
experimental conditions the ethyl ketone 8 did not react.
(7) Spanevello, R. A.; Pellegrinet, S. C. Synth. Comm. 1995, 25, 3663.
(8) In a typical procedure, the dienophile 1 (3.750 g, 13.59 mmol) was
dissolved in dry acetonitrile (136 mL) and stirred during 30 min with
anhydrous LiClO4 (7.2345 g, 67.95 mmol). Freshly cracked cyclopentadiene
(9.125 mL, 135.90 mmol) was added under an inert atmosphere at room
temperature. The reaction was completed after 18 h. The mixture was diluted
with ethyl ether and washed with water. The organic layer was dried over
Na2SO4 and concentrated under vacuum. Flash chromatography to remove
the excess of bicyclopentadiene afforded compound 2 (4.0480 g, 11.84
mmol) in 85% yield. The same results were obtained when the reaction
was catalyzed by BF3-OEt2 in methylene chloride at -78 °C or run under
thermal conditions in xylene at 80 °C, although in the later case the yield
was 51%. The 1H and 13C NMR signals of 2 were unequivocally assigned
by using homo- and heteronuclear 2D NMR techniques. The NOE observed
between C10-H and C13-H and between C12-H and C10-H (IUPAC
numbering) suggested the â approach. Further corroboration was obtained
from irradiation of C11-H and the carbonyl proton, which enhanced the
C8-H signal. The NOE observed between C11-H and one of the C15-H
showed the exo character of the addition.
A different structural modification was envisaged with the
purpose of gaining conformational flexibility. It implied the
cleavage of the benzylidene acetal ring and the subsequent
protection of the free hydroxyl groups as tert-butyl dimethyl
silyl ethers. Scheme 3 outlines the synthetic sequence
followed to obtain this derivative from known nitrile 9.7
In this case, the reaction of the dienophile 12 with
cyclopentadiene afforded a mixture of three different isomeric
products in 87% yield (Scheme 4). The exo-â cycloadduct
was again the main product, with a minor proportion of the
endo-â isomer and traces of a third adduct that could not be
purified from the mixture. The structure of the latter one
could not be unequivocally assigned, but as a result of the
(9) Pellegrinet, S. C.; Spanevello, R. A. Tetrahedron: Asymmetry 1997,
8, 1983.
(10) Pellegrinet, S. C.; Spanevello, R. A. Tetrahedron Lett. 1997, 38,
8623.
(13) All attempts to obtain these ketones directly from the aldehyde
precursor, the nitrile 9 (Scheme 3) through a nucleophilic addition were
fruitless.
(11) Rousch, W. R.; Brown, B. B. J. Org. Chem. 1992, 57, 3380.
(12) Takeda, K.; Imaoka, I.; Yoshii, E. Tetrahedron 1994, 50, 10839.
(14) The structure of the cycloadducts were assigned on the basis of the
NOE experiments as described in ref 8 for cycloadduct 2.
1074
Org. Lett., Vol. 2, No. 8, 2000

Table 1. Cycloaddition Reactions: Experimental Results and Spectroscopic Data
1H NMR chemical
13C NMR chemical
shift C-2/C-3 [ppm]
IR absorbance
obsd (theoretical) [cm-1
]
entry
dienophile
adducts [ratio]
yield [%]
shift vinyl proton [ppm]
1
2
1
12
exo-â
exo-â/endo-â/R
[27:6:1]
85
87
6.59
6.57
138.9/139.6
140.2/143.3
1690 (1695-1660)
1700 (1695-1660)
3
3
exo-â/endo-â
[2.2:1]
12
6.43
132.1/140.9
1696 (1675)
4
5
6
7
8
9
10
13
no reaction
no reaction
no reaction
exo-â
6.38
6.49
6.51
6.59
131.0/140.7
140.6/113.7
139.8/120.2
141.2/114.2
1702 (1675)
2227 (2240-2215)
2236 (2240-2215)
2230 (2240-2215)
15
difference in the coupling constant of the anomeric proton
in the 1H NMR spectrum, we assumed it was one of the two
possible isomers produced by the addition from the R face
of the molecule.
completely unreactive with cyclopentadiene under several
different conditions.15 Since it is known that a cyano group
has a withdrawing effect less strong than that of an
aldehyde,16 we removed the electron-rich benzylidene acetal
group and tested the nitrile 10, but it was also unreactive.
An alternative to increase the reactivity of the R,â-unsatur-
ated nitrile system was the protection of the free hydroxyl
groups with protecting groups that could have some electron-
withdrawing effect on the double bond. We found the acetate
Scheme 3
1
group could partially fulfill our expectation. The H NMR
spectrum of the diacetate 13 revealed a higher polarization
effect on the olefin. The vinyl proton signal was shifted
downfield, from 6.49 ppm in the bicyclic nitrile 9 to 6.58
ppm in the acetate derivative.
We further verified this effect experimentally by the
outcome of the reaction between nitrile 13 with cyclopen-
tadiene. The reaction afforded the exo-â isomer, as a single
product in 15% yield, plus recovered starting material. Table
1 summarizes the observed results.17
A comparative analysis of the spectroscopic data shows a
correlation with the observed results. The value of the
1
chemical shift of the vinyl proton in the H NMR spectra
gives a qualitative idea of the degree of polarization of the
double bond (hence, the activation for undergoing a cyclo-
addition process). A similar matter happens with the values
of the â olefinic carbon in the 13C NMR spectra. Usually
the â carbon of R,â-unsaturated systems suffers a deprotec-
tion effect due to this conjugation. Even though there is a
good correlation among the chemical shifts of protons and
carbons in both series of compounds, we consider it more
appropriate to compare the absolute value of the chemical
shift only within each series. Otherwise we could draw a
wrong conclusion by omitting to take into consideration other
effects.
A third type of modification was made on the electron-
withdrawing group. Replacement of the carbonyl group by
a cyano, which has a linear geometry, would avoid any
speculation on the dienophile s-cis or s-trans conformation.
However, the bicyclic nitrile 9 (Scheme 3) proved to be
Scheme 4
The IR data confirmed the decrease or lack of conjugation
of the R,â-unsaturated ketones, especially in the case of the
ethyl ketone 8. However, IR spectra of the nitrile series show
an absorbance peak (for the carbon-nitrogen stretching
(15) We tested the same conditions described for aldehyde 1 in ref 8
and also other Lewis acid catalysts: AlCl3,,MgBr2-OEt2, and TiCl4 in
methylene chloride at -78 °C.
(16) Sauer, J.; Wiest, H.; Mielert, A. Chem. Ber. 1964, 97, 3183.
(17) The spectroscopic and physical properties of all new compounds
reported herein are in complete agreement with the assigned structures.
Org. Lett., Vol. 2, No. 8, 2000
1075

vibrations) within the range for conjugated nitrile groups,
as was expected for a group with linear geometry.
In the case of the R,â-unsaturated nitrile series, it is known
that the activation effect of the cyano group is lower than
that of an aldehyde, and this could be responsible for the
lack of reactivity observed in entries 5 and 6. Only when an
additional electron-withdrawing group is present in the allylic
position (entry 7), is it possible to observe an increase in
the reaction rate. Moreover, since this reactivity is still very
low, it provokes a higher exo selectivity than in entry 2.
In view of the experimental evidences we could infer that
the observed stereoselectivity in the cycloaddition reaction
of aldehyde 1 is mainly caused by steric effects. When the
dienophile system becomes more flexible (entry 2 in Table
1) the selectivity decreases. This assumption is in agreement
with the steric hindrance effect caused by the methylene
group of the cyclopentadiene mentioned by Sauer.18 The
steric interaction between this methylene and the axial
hydrogen attached to C-4 and C-6 of the dienophile 1 are
probably responsible for the high exo selectivity. Although
the lower or lack of reactivity of the ketone derivatives
(entries 3 and 4) was attributed to a low degree of
polarization of the double bond, it is caused by a steric
interaction that does not allow the carbonyl group to adopt
the adequate conformation to attain a good conjugation with
the double bond. Entry 3 also showed the influence of the
carbonyl side chain on the exo selectivity.
Acknowledgment. Financial support from the Interna-
tional Foundation for Science (IFS), Third World Academy
of Sciences (TWAS), Universidad Nacional de Rosario,
Agencia Nacional de Promocio´n Cient´ıfica y Tecnolo´gica,
and CONICET is gratefully acknowledged. The authors
thank Prof. E. A. Ru´veda for helpful suggestions, and S.C.P.
thanks CONICET for the award of a fellowship.
Supporting Information Available: Characterization
data for compounds 1-5 and 8-17. This material is available
free of charge via the Internet at http://pubs.acs.org.
(18) Sauer, J.; Sustmann, R. Angew. Chem., Int. Ed. Engl. 1980, 19, 779
and references cited therein.
OL005619K
1076
Org. Lett., Vol. 2, No. 8, 2000