Diels-Alder reactions of chiral 1,3-dienes

Article abstract of DOI:10.1021/jo9609743

Structurally related 1,3-dienes 3a-f attached to a chiral carbon were treated with maleic anhydride under a variety of conditions. The initial Diels-Alder adducts were not isolated as they spontaneously rearranged to the isomeric isoindolones 6 and 7. The rate of the rearrangement was dependent on the relative stereochemistry of the initial cycloadducts. The π-facial selectivity of the Diels-Alder reaction was determined by analyzing reaction mixtures of the isomeric isoindolones by ¹H and ¹³C NMR spectroscopy and HPLC. A gradual increase in the π-facial selectivity was observed when the homoallylic hydroxyl group of the diene was endowed with a larger protecting group or slightly increased by performing the reactions in ether saturated with LiClO₄. These effects were rationalized by proposing a model based on 1,3-allylic strain in the transition state for the dienes 3a-f.

Full text of DOI:10.1021/jo9609743

J . Org. Chem. 1996, 61, 8425-8431
8425
Diels-Ald er Rea ction s of Ch ir a l 1,3-Dien es
Geoffrey T. Crisp* and Markus G. Gebauer
Department of Chemistry, The University of Adelaide, Adelaide 5005, South Australia
Received May 28, 1996X
Structurally related 1,3-dienes 3a -f attached to a chiral carbon were treated with maleic anhydride
under a variety of conditions. The initial Diels-Alder adducts were not isolated as they
spontaneously rearranged to the isomeric isoindolones 6 and 7. The rate of the rearrangement
was dependent on the relative stereochemistry of the initial cycloadducts. The π-facial selectivity
of the Diels-Alder reaction was determined by analyzing reaction mixtures of the isomeric
isoindolones by 1H and 13C NMR spectroscopy and HPLC. A gradual increase in the π-facial
selectivity was observed when the homoallylic hydroxyl group of the diene was endowed with a
larger protecting group or slightly increased by performing the reactions in ether saturated with
LiClO4. These effects were rationalized by proposing a model based on 1,3-allylic strain in the
transition state for the dienes 3a -f.
1,3-Dienes are important intermediates in organic
synthesis as they can be elaborated to six-membered
carbocycles via a Diels-Alder reaction and up to four
stereogenic centers can be formed in a single step.1 The
π-facial selectivity of 1,3-dienes that are attached to an
adjacent asymmetric carbon atom has been explored in
the synthesis of natural products,2 and several studies
have examined the parameters that govern this selectiv-
ity.3 A greater understanding of these factors should lead
to the design of 1,3-dienes that afford only a single
diastereomer with the desired stereochemistry. High
π-facial selectivity appears to be a feature of 1,3-dienes
that possess a substituent in the cis-position,4ab although
individual examples of high facial selectivity (>95:5) have
been reported for dienes lacking this substituent.4c The
steric interactions between the substituents on the
adjacent stereogenic center and the cis-substituent are
referred to as 1,3-allylic strain and must be minimized
in order to obtain high diastereoselectivities.5,6 Chiral
1,3-dienes that have a hydrogen atom in the cis-position
usually display only a modest π-facial selectivity.4a,b In
addition, both electronic and conformational effects can
influence the π-facial bias7 as well as the nature of the
dienophile.4c
It was the purpose of this study to investigate the effect
of varying the size of the homoallylic hydroxyl protecting
group and the allylic amino group on the π-facial selec-
tivity of a series of 1,3-dienes. Knowledge about these
factors is expected to be useful when chiral 1,3-dienes
are designed as intermediates in the synthesis of chiral
natural products.
Resu lts
The synthesis of dienes 3a -f from 2a -f and 1 is
described in previous papers (Figure 1).8,9 It was antici-
pated that a change to either the amino or hydroxyl
protecting groups of dienes such as 3 would provide a
means of varying the size of the substituents on the
stereogenic carbon and therefore assess their resulting
effect on the π-facial selectivity.
Diene 3a was reacted with maleic anhydride in 1,2-
dichlorobenzene (Figure 1, condition A). 1H NMR analy-
sis of the crude reaction mixture showed two sets of
resonances that were assigned to the diastereomeric
products 6a and 7a as described below. The IR spectrum
of the crude mixture was consistent with a carboxylic acid
and a tertiary cyclic imide. While isoindolone 6a was
isolated in 30% yield by HPLC, only a contaminated
sample of the minor product 7a could be obtained.
The reaction was then attempted using a concentrated
solution of diene 3a with maleic anhydride in refluxing
CH2Cl2 (Figure 1, condition B). 1H NMR analysis of the
crude reaction mixture revealed the same predominant
product 6a as for the reaction conducted at 180 °C. The
crude reaction mixture was extracted with saturated
NaHCO3, and the acidified extracts were purified by
column chromatography to yield carboxylic acid 6a in
47%. Maleic anhydride did not react with alkene 2a ,
which suggested that the Diels-Alder reaction precedes
the attack of the imine nitrogen on the anhydride, and
therefore, an intermolecular Diels-Alder reaction had
occurred and was followed by the rearrangement (Figure
1). It is also possible that a small amount of cycloadducts
6a and 7a formed initially and catalyzed the enolization
of the carbamate group.10
* To whom correspondence should be addressed. E-mail: gcrisp@
chemistry.adelaide.edu.au.
X Abstract published in Advance ACS Abstracts, November 1, 1996.
(1) Carruthers, W. Cycloaddition Reactions in Organic Synthesis;
Pergamon Press: New York, 1990.
(2) (a) Kozikowski, A. P.; Nieduzak, T. R.; Konoike, T.; Springer, J .
P. J . Am. Chem. Soc. 1987, 109, 5167-5175. (b) Grieco, P. A.; Beck, J .
P. Tetrahedron Lett. 1993, 34, 7367-7370.
(3) (a) Masamune, S.; Choy, W.; Petersen, J . S.; Sita, L. R. Angew.
Chem., Int. Ed. Engl. 1985, 24, 1-30. (b) Trost, B. M.; O’Krongly, D.;
Belletire, J . L. J . Am. Chem. Soc. 1980, 102, 7595-7596. (c) Stoodley,
R. J .; Gupta, R. C.; Slawin, A. M. Z.; Williams, D. J . J . Chem. Soc.,
Chem. Commun. 1986, 668-670. (d) Stoodley, R. J .; Gupta, R. C.;
Harland, P. A. J . Chem. Soc., Chem. Commun. 1983, 754-756. (e)
Bednarski, M; Danishefsky, S. J . Am. Chem. Soc. 1983, 105, 6968-
6969. (f) Ito, Y.; Amino, Y.; Nakatsuka, M.; Saegusa, T. J . Am. Chem.
Soc. 1983, 105, 1586-1590. (g) Tripathy, R.; Franck, R. W.; Onan, K.
D. J . Am. Chem. Soc. 1988, 110, 3257-3262.
(4) (a) McDougal, P. G.; Rico, J . G.; Van Derveer, D. J . Org. Chem.
1986, 51, 4492-4494. (b) Franck, R. W.; Argade, S.; Subramaniam, C.
S.; Frechet, D. M. Tetrahedron Lett. 1985, 26, 3187-3190. (c) McDou-
gal, P. G.; J ump, J . M.; Rojas, C.; Rico, J . G. Tetrahedron Lett. 1989,
30, 3897-3900.
(5) Adam, W.; Gla¨ser, J .; Peters, K.; Prein, M. J . Am. Chem. Soc.
1995, 117, 9190-9195.
(6) Hoffmann, R. W. Chem. Rev. 1989, 89, 1841-1860.
(7) Macaulay, J . B.; Fallis, A. G. J . Am. Chem. Soc. 1990, 112, 1136-
1144.
(8) Gebauer, M. G.; Crisp, G. T. Tetrahedron, in press.
(9) Crisp, G. T.; Glink, P. T. Tetrahedron 1992, 48, 3541-3556.
(10) An acid-catalyzed rearrangement of a similar system was
reported in ref 2b, and so a small amount of 6a or 7a could have
catalyzed the cyclization.
S0022-3263(96)00974-7 CCC: $12.00 © 1996 American Chemical Society

8426 J . Org. Chem., Vol. 61, No. 24, 1996
Crisp and Gebauer
Ta ble 1. Th e Syn th esis of Isoin d olon es 6 a n d 7 via th e
Diels-Ald er Rea ction of Dien es 3 w ith Ma leic An h yd r id e
amino
protecting protecting
group
hydroxyl
yield yield
of 6 of 7 deb
entry diene
group
COCH3
COCH3
COCH3
COCH3
none
COCMe3
COCMe3
COCMe3
SiMe2-t-hexyl
SiMe2-t-hexyl
SiMe2-t-hexyl
SiPh2-t-Bu
SiPh2-t-Bu
SiPh2-t-Bu
SiPh2-t-Bu
COCH3
condnsa (%) (%) (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
3a
3a
3a
3a
3b
3c
3c
3c
3d
3d
3d
3e
3e
3e
3e
3f
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Cbz
Boc
A
B
C
D
C
A
B
C
A
B
C
A
B
C
D
C
30
47
38
42
ndd
36
55
44
34
62
46
39
63
48
49
18
7c 46
ndd
56
69
ndd
54
9
ndd
65
15c 63
ndd
74
11 64
ndd
76
83
57
a
Reactions were performed under the conditions listed in Figure
b
1. de ) diastereomeric excess of 6 over 7 determined by NMR
analysis of crude reaction mixtures. c Only partially purified. nd
d
) not determined.
performed. The ratio of isoindolone 7a and its precursor
5a with regard to isoindolone 6a was constant. This
constant relationship is consistent with a slow rearrange-
ment of 5a (which could not be isolated) as it accumulated
during the reaction, while the isomeric 4a rearranged too
rapidly to be detected. The constant ratio of the peak
areas was interpreted as the underlying π-facial selectiv-
ity of diene 3a during the Diels-Alder reaction (de 56%).
A direct interconversion of isoindolones 6a and 7a was
ruled out by heating isomerically pure 6a at 180 °C in
1,2-dichlorobenzene. After the solvent was distilled, only
1
unchanged 6a was detected by H NMR analysis of the
crude material. This experiment also suggested that the
Diels-Alder reaction was under kinetic and not thermo-
dynamic control.
F igu r e 1. Diels-Alder reactions of dienes 3a -f with maleic
anhydride.
1
To analyze the reaction mixture directly by H NMR,
When the Diels-Alder reactions of other dienes (3c-
e) bearing a Cbz amino protecting group but different
hydroxyl protecting groups were investigated (Figure 1),
a similar rate difference in the rearrangement of the
intermediate cycloadducts 4c-e and 5c-e was observed.
In contrast, both Boc-protected initial adducts 4f and 5f
rearranged rapidly, which was evidenced by a ratio of
isomeric isoindolones 6g and 7g which remained constant
during the reaction.
With the exception of alcohol 6b all isoindolones 6 were
isolated isomerically pure by chromatography of the
crude Diels-Alder products in yields ranging from 18 to
63% (Table 1). Although, on the basis of a 1H and 13C
NMR analysis of the reaction mixtures, isoindolone 6b
was formed when diene 3b was treated with maleic
anhydride in CDCl3 at 25 °C, the isolation of 6b was
thwarted by the presence of significant amounts of
unidentified material. A pure reference sample of alcohol
6b was synthesized by the acidic hydrolysis of the acetyl-
protected isoindolone 6a in 96% yield.
For the isolation of the minor Diels-Alder products 7,
dienes were reacted with maleic anhydride in refluxing
1,2-dichlorobenzene to increase the relative amount of 7
in the crude product. Among the isoindolones 7, only the
pivaloyl, and tert-butyldiphenylsilyl-protected adducts 7c
and 7e were isolated pure in very modest yields of 9 and
11%, respectively. The acetyl- and the tert-hexyldimeth-
ylsilyl-protected isoindolones 7a and 7e were only par-
tially purified due to the presence of unidentified impu-
the reaction of diene 3a with maleic anhydride was
conducted in CDCl3 at 25 oC (Figure 1, condition C).
Although isoindolone 6a was the major product as before,
the ratio of isomeric isoindolones 6a and 7a present in
the crude product varied. It appeared that extended
reaction times led to a decrease in the ratio of 6a to 7a .
If the rearrangement of the cycloadducts 4a and 5a
proceeded at distinctly different rates, then isoindolone
6a would predominate due to the rapid rearrangement
of the initial cycloadduct 4a and a slow rearrangement
of cycloadduct 5a with time would lead to formation of
7a .
To confirm this the Diels-Alder reaction of diene 3a
with maleic anhydride was monitored by HPLC (Figure
1, condition C). The initial ratio of 6a to 7a of 33.5:1.0
at 12 h was gradually reduced to 13.8:1.0 after 4 days. A
late-eluting peak due to a nonpolar compound was also
detected and decreased in intensity with time. When the
reaction mixture was analyzed by 1H NMR additional
signals remained transiently after the starting diene 3a
was completely consumed. Presumably, a reaction in-
termediate such as the slowly rearranging cycloadduct
5a had accumulated and gave rise to the late eluting peak
in the HPLC traces. Refluxing the crude reaction product
in 1,2-dichlorobenzene at 180 °C led to the disappearance
of the late eluting peak. A quantitative analysis of the
ratio of the sum of the peak areas in the chromatograms
obtained after varying reaction times at 25 °C was

Diels-Alder Reactions of Chiral 1,3-Dienes
J . Org. Chem., Vol. 61, No. 24, 1996 8427
tical in all isoindolones. The assignment rested on NOE
and decoupling difference spectroscopy.14
On the basis of Alder’s endo rule,15 both isoindolones
6a and 7a were expected to be derived from endo
cycloadducts that only differed in the relative stereo-
chemistry at C1 and C9a. Molecular models showed that
only for the trans-isoindolones 6 was the dihedral angle
spanned by H1 and H9a close to 90°. This suggested that
the major product 6a had the trans-stereochemistry
depicted in Figure 1, while the minor product 7a had the
cis-stereochemistry (the terms “trans” and “cis” refer to
protons H1 and H9a being on either opposite faces or on
the same face of the five-membered ring). Unequivocal
evidence for the stereochemical assignment of 7a was
obtained by the X-ray diffraction analysis of a single
crystal of the methyl ester derivative 7g,16 which was
obtained when the crude product of the reaction of diene
3a with maleic anhydride in 1,2-dichlorobenzene (Figure
1) was esterified under neutral conditions with diazo-
methane in a yield of 74%. During the esterification no
selective kinetic effects were observed since the two
methyl esters 6g and 7g were produced in the same ratio
as the parent carboxylic acids 6a and 7a (1H NMR and
HPLC). When a sample of the crude esters 6g and 7g
was dissolved in a minimum amount of MeOH, crystals
of 7g precipitated. Purification of the mother liquor by
HPLC resulted in the isolation of methyl ester 6g in 21%
yield. 1H NMR analysis of this material confirmed that
it was identical to the major compound present in the
crude product of the esterification reaction.
The relative stereochemistry of ester 6g was ascer-
tained by steady-state NOE difference spectroscopy.14,17
Selective irradiation of the diastereotopic protons in
position 10 produced small but measurable NOE en-
hancements of less than 1% for H9a and for H3a. These
NOEs suggested that the acetyl group and H9a and H3a
were on the same face of the five-membered ring, which
was consistent with the assignment of the trans-endo-
stereochemistry to 6g. NOE enhancements of less than
1% that were complementary to those measured in the
previous experiment were observed for H10 when H3a
or H9a of 6g were selectively irradiated. The absolute
stereochemistry of isoindolones 6a , 7a , 6c, and 7 was
predetermined by the (S) configuration of C2 in dienes
3. The stereochemical assignment was extended to all
isoindolones, which had very similar 1H and 13C NMR
spectra.
F igu r e 2. Stereochemistry of 6a .
rities with identical retention times during chromatog-
raphy on normal and reversed-phase silica.
The most remarkable structural feature of the isoin-
dolones depicted in Figure 1 was their rigid cyclic
structure, which was reflected in the appearance of the
1H and 13C NMR spectra. Isoindolones having the same
relative stereochemistry displayed a great similarity in
the 1H and 13C NMR data, as well as H,H COSY and H,C
COSY spectra.
An interesting feature in the 1H NMR spectra of
isoindolones 6 was a small (J ) 2 Hz) homoallylic proton
coupling between H9a and H4a as well as H9a and H8ax,
which indicated a rigid collinear alignment of the H9a-
C9a, H8ax-C8, and H4a-C4a proton carbon bonds (Fig-
ure 2).11 When the dihedral angle is near 90° no vicinal
coupling may be observed.12 This was indeed the case
for the vicinal protons H1 and H9a of isoindolones 6 as
evidenced by selective homonuclear decoupling experi-
ments (detection limit: J ) 0.5 Hz). On the contrary, a
coupling of J ) 7 Hz was measured for the same pair of
protons (H1 and H9a) in the isomeric isoindolone 7, which
indicated a different relative stereochemistry.
As the connectivity between positions 1 and 9a of 6
could not be established via proton proton coupling, a
COLOC experiment13 was conducted. Cross peaks were
observed between H9 and carbons in positions 1, 3a, 4a,
8, and 9a (Figure 2). Furthermore, H1 and C9a were
correlated as well as one of the diastereotopic protons in
position 10 with C9a. These results provided unambig-
ous proof of the connectivity between positions 1 and 9a
of the methyl ester derivative 6g for which no vicinal
proton coupling was observed and allowed for a complete
assignment of the 1H and 13C NMR spectra of 6. The
appearance of protons in positions 5, 6, 7, and 8 was
independent of the stereochemistry and practically iden-
The complete conversion of the starting dienes (3a -f)
was ensured by refluxing the crude products in 1,2-
dichlorobenzene for 1 h. Thereafter, the ratios of trans-
to cis-isoindolones were interpreted as a measure for the
π-facial selectivity of the respective dienes. As a single
exception the de of the Diels-Alder reaction of the
unprotected diene 3b was not determined since the
reaction mixture decomposed significantly.
In the case of the acetyl- and pivaloyl-protected isoin-
1
dolones 6c and 7c this ratio was measured by H NMR
spectroscopy. This method of analysis could not be
extended to silyl ether-protected isoindolones 6d and 7d
(15) (a) Alder, K.; Schumacher, M.; Wolff, O. Liebigs Ann. Chem.
1949, 564, 79-96. (b)Alder, K. Ann. 1951, 571, 157-166.
(16) Gebauer, M. G.; Crisp, G. T.; Tiekink, E. R. T. Z. Kristallogr
1995, 210, 465-466.
(17) Neuhaus, D. The Nuclear Overhauser Effect in Structural and
Conformational Analysis; VCH Verlag: Weinheim, 1989.
(18) (a) Abraham R. J . Introduction to NMR Spectroscopy; Chi-
chester: Wiley, 1988. (b) Sanders, J . K. M.; Hunter, B. K. Modern NMR
Spectroscopy, 2nd ed.; Oxford University Press: Oxford, 1993.
(11) Williams, D. H.; Fleming, I. Spectroscopic Methods in Organic
Chemistry, 4th ed.; McGraw Hill: London, 1989.
(12) Karplus, M. J . Am. Chem. Soc. 1963, 85, 2870-2871.
(13) Kessler, H.; Griesinger, C.; Zimmermann, G. J . Magn. Resson.
1984, 57, 331-336.
(14) Sanders, J . M.; Marsh, J . D. Prog. Nuclear Magn. Reson.
Spectrosc. 1982, 15, 353-400.

8428 J . Org. Chem., Vol. 61, No. 24, 1996
Crisp and Gebauer
as the corresponding proton signals, as well as those of
6e and 7e, were overlapping, and HPLC analysis proved
unsuccessful. However, the tertiary olefinic resonances
of 6d and 7d were well separated by approximately 4
ppm in both cases.
Several trends concerning the π-facial selectivity of
dienes in the Diels-Alder reaction with maleic anhydride
can be seen in Table 1. Larger hydroxyl protecting
groups led to a gradual increase in the π-facial selectivity
of the Diels-Alder reaction. In particular, the acetyl-
protected diene 3a had a lower π-facial selectivity than
the pivaloyl-protected diene 3c (Table 1, entries 3 and
8). A plateau of de ∼75% was reached with the very
bulky silyl ethers 3d and 3e (Table 1, entries 11 and 14).
These results clearly implicate the hydroxyl protecting
group in the steric shielding of the 1,3-diene moiety. The
nature of the alkyl portion of the carbamate protecting
group appeared to have no influence on the π-facial bias
of the investigated dienes, suggesting that the carbamate
group was directed away from the site of reaction in the
transition state (Table 1, entries 3 and 16). The π-facial
selectivity of the investigated dienes deteriorated with
an increase in the reaction temperature from ambient
temperature to 180 °C, as is expected for a reaction
conducted under kinetic control (entries 1 and 3). Nev-
ertheless, a de of 64% for the reaction of the siloxy-
protected diene 3e indicated that the π-facial selectivity
was still relatively high even at the elevated temperature
(Table 1, entry 12). A change in the reaction solvent from
CDCl3 to a saturated ethereal solution of LiClO4 led to a
slight increase in the π-facial selectivity (Table 1, entries
14 and 15). Such a solvent effect has been previously
reported.19
F igu r e 3.
that the eclipsed conformations of the carbon-carbon
double bond with a substituent on the allylic carbon are
generally lower in energy than the alternative staggered
arrangements. Applying the literature results to dienes
3a -f gives three ground-state conformations of low
energy (Figure 3) by rotating the C2-C3 bond of dienes
3. It is interesting to note that conformation 9 is
essentially the solid-state conformation of diene 3a .23
Assuming that the same order of stability is main-
tained in the transition state conformation of dienes 3
depicted in Figure 3,6 minimizing 1,3-allylic strain pre-
dicts the correct stereochemistry for the major Diels-
Alder products 6. In the predominant conformer 10 in
which 1,2- and 1,3-eclipsing interactions are minimized,
one face of the planar diene unit was shielded against
the approach of maleic anhydride by the hydroxyl pro-
tecting group, while the other face could be readily
accessed leading to the formation of the major trans-
isoindolone 6. Since dienes 3a and 3f displayed the same
π-facial selectivity (Table 1, entries 3 and 16) the amino
protecting group was directed away from the diene unit.
Furthermore, studies on CPK molecular models of diene
3a indicated that the alkyl portion of the carbamate
protecting group could not be located in the proximity of
the diene unit without changing the hybridization of the
imide nitrogen from sp2 to sp3 and introducing consider-
able strain into the molecule.
Discu ssion
Many publications on the Diels-Alder reaction of
acyclic chiral dienes have appeared in the literature,3-5,7
and several explanations for their π-facial selectivity have
been advanced. Essentially all rationales focused on an
identification of the factors that stabilize the conforma-
tions that are accessed by rotation of the allylic C2-C3
bond as this was expected to be a key determinant for
the π-facial selectivity.
Fallis7 proposed that conformations that allow for a
stabilization of the emerging σ*-orbital on the carbon
adjacent to the allylic carbon (for example, C3 in diene
3a ) by hyperconjugation from a collinear carbon-carbon
or carbon-hydrogen bond (C1-C2 or C2-H2 in diene 3a )
would be preferred. Stereoelectronic effects of the latter
kind are referred to as the “Cieplak-effect”.20 For some
substituted acyclic 1,3-dienes a different model was
proposed by Prein.5 These dienes are distinguished from
those with a hydrogen at C4 by a ground-state preference
for conformations that minimize 1,3-allylic strain. Argu-
ably, the transition state conformation of these dienes
in the Diels-Alder reaction is also subject to conforma-
tional control due to allylic strain.6 NMR evidence21 and
recent ab initio molecular orbital calculations22 revealed
Diene conformation 9 would then give the minor cis-
isoindolone 7. Diene conformation 8, on the other hand,
was presumably least stable and therefore contributed
only in a negligible way to the transition state, since
significant nonbonding interactions between the CH2OPg
moiety and H4 existed.
Most importantly, this model based on the minimiza-
tion of 1,3-allylic strain5 and minimization of 1,2-eclipsing
strain accounted for the observed increase in the π-facial
selectivity with an increase in the size of the hydroxyl
(19) (a) Aversa, M. C.; Bonaccorsi, P.; Giannetto, P. Tetrahedron
Asymmetry 1994, 5, 805-808. (b) Casaschi, A.; Desimoni, G.; Faita,
G.; Invernizzi, A. G.; Lanati, S.; Righetti, P. P. J . Am. Chem. Soc. 1993,
115, 8002-8007.
(20) (a) Cieplak, A. S. J . Am. Chem. Soc. 1981, 103, 4540-4552. (b)
Cieplak, A. S.; Tait, B. D.; J ohnson, C. R. J . Am. Chem. Soc. 1989,
111, 84478462.
(22) Brocker, J . L.; Hoffmann, R. W.; Houk, K. N. J . Am. Chem.
Soc. 1991, 113, 5006-5017.
(23) Gebauer, M. G.; Crisp, G. T.; Tiekink, E. R. T. Z. Kristallogr.
1995, 210, 623-624.
(21) Gung, B. W.; Melnick, J . P.; Wolf, M. A.; King, A. J . Org. Chem.
1995, 60, 1947-1951.

Diels-Alder Reactions of Chiral 1,3-Dienes
J . Org. Chem., Vol. 61, No. 24, 1996 8429
optimized for J ) 8 Hz. H,H COSY25 and H,C COSY26
experiments were performed in a routine fashion and the data
multiplied with a shifted sine-bell window.
protecting group of dienes 3. Small hydroxyl protecting
groups, such as the acetyl of diene 3a , could easily be in
positions removed from the diene unit by a rotation
around the C1-C2 bond. The hydrogen atoms in position
1 were too small to effectively shield the diene unit from
an incoming dienophile in conformer 10. However, even
for the silyl ether-protected dienes 3d and 3e a plateau
of de ∼75% was approached as both diene conformations
9 and 10 contributed to the transition state. The
alternative proposal based on stereoelectronic effects
could not explain the trends observed in Table 1.7,24
When the Diels-Alder reaction of dienes 3a and 3e
with maleic anhydride was conducted in ether saturated
with LiClO4, coordination and activation of the dienophile
by lithium ions resulted and the effective size of the
dienophile was increased. Conceivably, such an increase
led to a better π-facial discrimination of conformation 10,
as the hydroxyl protecting group provided better shield-
ing of the 1,3-diene moiety against the approach of a
comparatively larger lithium-coordinated dienophile.
Accurate isomeric ratios by 13C NMR were obtained by
minimizing NOE effects.18 Proton decoupling was applied only
during the acquisition time, and chromium acetylacetonate
was added to suppress NOEs from protons onto carbon spins
and to accelerate the relaxation of carbon spins. Under these
conditions, the T1 relaxation times of the tertiary olefinic 13C
spins were measured to be less than 200 ms, and a delay period
of 2 s between scans was deemed adequate for a thermal
equilibrium to re-establish.
(1S,3a S,4R,4a R,9a S)-1-(Acetoxym eth yl)-2-N-(ben zoxy-
car bon yl)-3-oxo-3a,4,4a,5,6,7,8,9a-octaah ydr oben z[f]isoin -
d ole-4-ca r boxylic Acid (6a ) (Figure 1, Condition B).
A
mixture of diene 3a (600 mg, 1.747 mmol), activated 4A
molecular sieves (707 mg), and maleic anhydride (514 mg,
5.241 mmol) in CH2Cl2 (5 mL) was refluxed for 1 d. A small
aliquot of the reaction mixture was withdrawn, the solvent
evaporated, and the residue analyzed by 1H and 13C NMR
spectroscopy. Neither starting diene 3a nor cis-isoindolone 7a
was detected, and resonances due to trans-isoindolone 6a
dominated the spectra. The reaction mixture was filtered, the
molecular sieves were washed with CH2Cl2, the solvent was
evaporated from the combined organic phases, and the residue
was dissolved in ether (100 mL). It was extracted with ice-
cold saturated NaHCO3 (3 × 30 mL), and the combined
extracts were immediately washed with ether (30 mL) and
acidified by slowly adding concentrated HCl to pH ) 0. The
mixture was extracted with ether (3 × 50 mL), the combined
extracts were dried (MgSO4), and the was solvent evaporated
to yield an amorphous light brown solid (531 mg, 69%). The
crude product was analyzed by HPLC, and the following peaks
were observed (relative intensities): t ) 20.9 min (6b, 14%), t
) 32.3 min (unknown, 8%), t ) 33.8 min (7a , 2.5%), t ) 35.3
min (unknown, 11%), t ) 40.5 min (6a , 100%). The crude
product was chromatographed on silica (120:120:12.5 hexane:
Con clu sion s
This study provided evidence that the π-facial selectiv-
ity of 1,3-dienes that have an attached stereogenic center
can be enhanced (up to de ) 83%) by increasing the size
of the homoallylic protecting hydroxyl group in dienes 3
and conducting the reaction in LiClO4/ether. A confor-
mational model based on 1,3-allylic strain was proposed
to rationalize the stereochemical outcome of the Diels-
Alder reactions and the gradual increase in the π-facial
selectivity of dienes 3 with progressively larger hydroxyl
protecting groups.
CH2Cl2:AcOH) to yield 6a (47%) as an oil: [R]25 ) -61.8
D
(c2.72, CDCl3); IR (neat) 3300-2400, 1785, 1750-1730 cm-1
;
Exp er im en ta l Section
1H NMR δ 13.2 (bs, 1 H), 7.42-7.33 (m, 5 H), 5.31, 5.26 (AB,
J ) 12 Hz, 2 H, Bn), 5.18 (bs, 1 H, H9), 4.43 (dd, J ) 5, 12 Hz,
1 H, H10), 4.21 (dd, J ) 4, 12 Hz, 1 H, H10), 4.13 (m, 1 H,
H1), 3.65 (dd, J ) 4, 8 Hz, 1 H, H3a), 3.01 (m, 1 H, H4), 2.91
(m, 1 H, H9a), 2.44 (m, 1 H, H4a), 2.23 (b, 1 H, H8eq), 2.07 (s,
3 H, CH3), 2.0 (m, 2 H, H5eq, H8ax), 1.8 (bs, 2 H, H6eq, H7eq),
1.4 (m, 1 H, H7ax), 1.3 (m, 2 H, H5ax, H6ax); 13C NMR δ 175.6,
175.0, 170.5, 151.1, 146.2 (C8a), 134.9, 128.6, 128.5, 127.9 (Ar),
116.6 (C9), 68.6 (Bn), 63.9 (C10), 61.6 (C1), 43.2 (C4), 40.4
(C3a), 38.6 (C4a), 36.6 (C8), 36.4 (C9a), 31.6 (C5), 29.0 (C7),
26.9 (C6), 20.8 (CH3). Anal. Calcd for C24H27NO7: C, 65.29;
H, 6.16; N, 3.17. Found: C, 64.97; H, 6.40; N, 2.96.
The synthesis of alkenes 2 and dienes 3 is reported
separately.8 Contaminating acid was removed from maleic
anhydride by recrystallization (CHCl3) and sublimation. All
experiments were conducted under an atmosphere of nitrogen.
HPLC was carried out on a reversed-phase C18-silica column
(25 cm × 10 mm, particle size: 5 mm) at a flow rate of 4 mL/
min using a 355:520:125:3 mixture of CH3CN, H2O, MeOH,
and trifluoroacetic acid for the carboxylic acids or a 7:3 mixture
of MeOH and H2O for the methyl esters. Semipreparative
HPLC was carried out under the same conditions using less
H2O in the solvent systems. The eluant was generally detected
at 258 nm (Cbz derivatives) or at 240 nm (Boc derivatives).
All NMR samples were analyzed as solutions in CDCl3 at a
static field strength of 300 MHz (for 1H). For quantitative
determination of carbon signals a concentrated sample (>70
mg) was analyzed in the presence of a few milligram of
chromium(III) acetylacetonate. The pulse angle was 90° and
the acquisition time 4 s. A radiationless delay of 2 s was
included between scans. No window function was applied to
the free induction decay signal. For NOE difference spectros-
copy proton resonances were selectively irradiated for 0.8 s
before each scan using a continous wave of weak power (5-
25 mW) on the decoupler channel. A radiationless delay of 2
s was included. Typical parameters for COLOC experiments13
were in the t1-dimension (1H) (spectral width ) 2300 Hz;
number of experiments ) 180; number of data points ) 512);
in the t2-dimension (13C) (spectral width ) 18 000 Hz; acquisi-
tion time ) 0.17 s). The delay before the data acquisition was
(Figure 1, Condition D). A mixture of diene 3a (62 mg, 0.181
mmol), LiClO4 (213 mg, 2.00 mmol), and maleic anhydride (53
mg, 0.544 mmol) in dry ether (0.40 mL) was stirred in a sealed
vessel at rt for 5 h until none of the starting diene 3a was
detected by TLC analysis (2:3 EtOAc:hexane). Ether (8 mL)
and 1 M HCl (3 mL) were added, and the mixture was
thoroughly shaken. The separated aqueous phase was ex-
tracted with ether (2 × 5 mL), the combined organic phases
were washed with 1 M HCl (2 × 3 mL) and dried (MgSO4),
and the solvent was evaporated. The crude product was
dissolved in 1,2-dichlorobenzene (5 mL) and heated at 180 °C
for 1 h. Following distillation of the solvent (7 Torr), the ratio
of 6a :7a was 5.5 to 1.0 (1H NMR). Chromatography of the
crude reaction product afforded pure 6a in 42% yield.
(1S,3a S,4R,4a R,9a S)-1-(Hyd r oxym eth yl)-2-N-(ben zoxy-
car bon yl)-3-oxo-3a,4,4a,5,6,7,8,9a-octah ydr oben zo[f]isoin -
d ole-4-ca r boxylic Acid (6b) (Figure 1, Condition C). Diene
3b (102 mg, 0.339 mmol) was treated with maleic anhydride
(100 mg, 1.017 mmol) in CDCl3 (0.5 mL) at rt for 24 h.
Analysis of the crude reaction mixture by 1H and 13C NMR
spectroscopy revealed the presence of 6b as the major com-
(24) (a) Cherest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968,
10, 21992204. (b) Anh, N. T. Tetrahedron 1973, 29, 3227-3232.
(25) (a) Aue, W. P.; Bartholdi, E.; Ernst, R. R. J . Chem. Phys. 1976,
64, 2229-2246. (b) Kessler, H.; Gehrke, M.; Griesinger, C. Angew.
Chem., Int. Ed. Engl. 1988, 27, 490-536. (c) Ernst, R. R.; Bodenhausen,
G.; Wakaun, A. Principles of Nuclear Magnetic Resonance in One and
Two Dimensions; Claredon: Oxford, 1987.
(26) Bax, A.; Morris, G. J . Magn. Reson. 1981, 42, 501-505.

8430 J . Org. Chem., Vol. 61, No. 24, 1996
Crisp and Gebauer
pound; however, other unidentified resonances were prominent
in both spectra. An isolation of the major product 6b was not
attempted.
for 6c. After the distillation of 1,2-dichlorobenzene the
remaining crude product and chromium(III) acetylacetonate
(4 mg, 0.011 mmol) were dissolved in CDCl3 (0.5 mL). The T1
relaxation times of the resonances at δ )117.4 ppm (6d ) and
at δ ) 113.0 ppm (7d ) were estimated by the inversion recovery
F r om 6a . Isoindolone 6a (56 mg, 0.127 mmol) was dis-
solved in a mixture of methanol (6 mL), water (3 mL), and
trifluoroacetic acid (1 mL) and stirred at rt for 10 h. The
solvent was evaporated and the residue dissolved in CH2Cl2
(10 mL). The organic phase was washed with 1 M HCl (3 mL)
and dried (MgSO4) and the solvent evaporated to afford 6b
(45 mg) as a colorless oil in 96% yield: 1H NMR δ 9.3 (bs, 2
H), 7.36 (m, 5 H), 5.32, 5.27 (AB, J ) 12 Hz, 2 H, Bn), 5.18
(bs, 1 H, H9), 4.04-3.68 (m, 3 H, H10, H1), 3.80 (m, 1 H, H3a),
3.10 (dd, 1 H, H4), 2.98 (m, 1 H, H9a), 2.46 (m, 1 H, H4a),
2.22 (m, 1 H, H8eq), 2.0 (m, 2 H, H5eq, H8ax), 1.8 (bs, 2 H, H6eq,
H7eq), 1.4 (m, 1 H, H7ax), 1.3 (m, 2 H, H5ax, H6ax); 13C NMR δ
150.8, 145.9 (C8a), 117.4 (C9), 68.3 (Bn), 64.7 (C10), 62.8 (C1),
43.2 (C4), 40.6 (C3a), 38.5 (C4a), 36.4 (C8), 36.2 (C9a), 31.5
(C5), 28.9 (C7), 26.8 (C6). Anal. Calcd for C22H25NO6: C,
71.91; H, 6.86; N, 3.81. Found: C, 71.77; H, 6.67; N, 3.60.
(1S,3a S,4R,4a R,9a S)-1-[(2′,2′-Dim eth yleth oxy)m eth yl]-
2-N-(ben zoxyca r bon yl)-3-oxo-3a ,4,4a , 5,6,7,8,9a -octa h y-
d r oben z[f]isoin d ole-4-ca r boxylic Acid (6c) (Figure 1, Con-
dition C). A solution of diene 3c (98 mg, 0.254 mmol) and
maleic anhydride (75 mg, 0.763 mmol) in CDCl3 (0.4 mL) was
stirred at rt for 24 h. The solvent was evaporated and the
residue dissolved in 1,2-dichlorobenzene (9 mL) and heated
to reflux. The solvent was distilled (8 Torr) and the residue
method.
A
13C NMR spectrum was acquired using inverse
1
gated H decoupling, and a ratio of 6.8:1.0 was determined for
these resonances in favor of 6d . Chromatography (140:100:
12.5 hexane:CH2Cl2:AcOH) of the crude reaction product
afforded 6d as a light red oil in 46%.
(1S,3a S,4R ,4a R ,9a S)-1-[(Dip h e n yl-t er t -b u t ylsiloxy)-
m eth yl]-2-N-(ben zoxycar bon yl)-3-oxo-1,3a,4,4a, 5,6,7,8,9a-
octa h yd r oben z[f]isoin d ole-4-ca r boxylic Acid (6e). (Fig-
ure 1,Condition B). Diene 3e (237 mg, 0.440 mmol) was
treated with maleic anhydride (129 mg, 1.319 mmol) in CH2-
Cl2 as described previously for 6c. The crude product was
chromatographed (140:100:12.5 hexane:CH2Cl2:AcOH) and 6e
obtained as a colorless oil in 63% yield: IR (neat) 3300-2500,
1
1780, 1735, 1700 cm-1; H NMR δ 13.6 (bs, 1 H), 5.24 (dd, 2
H,), 5.19 (bs, 1 H), 3.91 (m, 2 H, H10), 3.83 (m, 1 H, H1), 3.79
(m, 1 H, H3a), 3.02 (m, 1 H, H4), 2.97 (m, 1 H, H9a), 2.45 (m,
1 H, H4a), 2.21 (m, 1 H, H8eq), 2.0 (m, 2 H, H5eq, H8ax), 1.8
(bs, 2 H, H6eq, H7eq), 1.35-1.05 (m, 3 H, H7ax, H5ax, H6ax), 1.07
(s, 9 H, CH3); 13C NMR δ 179.1, 173.7, 150.6, 145.6, 135.5,
135,4, 134.6, 132.5, 132.2, 130.2, 130.1, 128.7, 128.6, 128.0 (Ar),
117.2 (C9), 68.6 (Bn), 64.9 (C10), 64.0 (C1), 45.5 (C4), 41.0
(C3a), 39.1 (C4a), 36.9 (C8), 36.4 (C9a), 31.6 (C5), 29.7, 28.8
(C7), 26.8 (CH3), 26.7 (C6). Anal. Calcd for C38H43NSiO6: C,
71.56; H, 6.80; N, 2.20. Found: C, 71.47; H, 6.88; N, 2.19.
(Figure 1, Condition C). Diene 3e (175 mg, 0.324 mmol)
was treated with maleic anhydride (96 mg, 0.974 mmol) in
CDCl3 as for 6c. Chromatography (140:100:12.5 hexane:CH2-
Cl2:AcOH) of the crude reaction product afforded 6e in 48%
yield as a light red oil.
1
analyzed by H NMR spectroscopy. A ratio of 4.65 to 1.0 for
the proton resonances at δ ) 3.71 ppm (6c) and at δ ) 3.40
ppm (7c) was measured. Chromatography (120:120:12.5 hex-
ane:CH2Cl2:AcOH) of the crude reaction product afforded 6c
in 44% yield as a colorless oil: 1H NMR δ 12.9 (bs, 1 H), 5.33
(s, 2 H, Bn), 5.19 (bs, 1 H, H9), 4.46 (m, 1 H, H10), 4.20 (m, 1
H, H10), 4.16 (m, 1 H, H1), 3.60 (m, 1 H, H3a), 3.12 (dd, J )
4, 7 Hz, 1 H, H4), 2.90 (m, 1 H, H9a), 2.49 (m, 1 H, H4a), 2.23
(b, 1 H, H8eq), 2.0 (m, 2 H, H5eq, H8ax), 1.8 (bs, 2 H, H6eq, H7eq),
1.4 (m, 1 H, H7ax), 1.2 (m, 2 H, H5ax, H6ax), 1.17 (s, 9 H, CH3);
13C NMR δ 177.7, 177.5, 173.3, 150.6, 146.3 (C8a), 134.7, 128.7,
128.7, 128.0 (Ar), 116.5 (C9), 68.9 (Bn), 63.7 (C10), 61.2 (C1),
44.7 (C4), 40.6 (C3a), 39.0 (C4a), 38.8 (quarternary), 36.8 (C8),
36.5 (C9a), 31.6 (C5), 28.9 (C7), 27.1 (CH3), 26.8 (C6). Anal.
Calcd for C27H33NO7: C, 67.06; H, 6.88; N, 2.90. Found: C,
66.72; H, 6.63; N, 2.59.
(Figure 1, Condition D). Diene 3e (156 mg, 0.289 mmol)
was treated with maleic anhydride (85 mg, 0.867 mmol) in
ether saturated with LiClO4 as for 6a . After 1,2-dichloroben-
zene was distilled, the remaining crude product was analyzed
by 13C NMR spectroscopy in the presence of chromium(III)
acetylacetonate as outlined for 6d and showed a ratio of 11.0
to 1.0 for 6e and 7e. Chromatography (140:100:12.5 hexane:
CH2Cl2:AcOH) of the crude reaction product afforded 6e in 49%
yield as a light red oil.
(Figure 1, Condition B). Diene 3c (183 mg, 0.475 mmol)
was treated with maleic anhydride (140 mg, 1.426 mmol) in
refluxing CH2Cl2 (2.0 mL) in the presence of activated A4
molecular sieves (200 mg). After 12 h only a trace of diene 3c
was detected by TLC analysis of the reaction mixture (3:7
EtOAc:hexane), which was then cooled to rt. The molecular
sieves were removed and washed with CH2Cl2. The combined
organic phases were evaporated and the crude product chro-
matographed to yield 6c (55%).
(1S,3a S,4R,4a R,9a S)-1-[[Dim eth yl(1′,1′,2′-tr im eth ylp r o-
p y l)s i lo x y ]m e t h y l]-2-N -(b e n z o x y c a r b o n y l)-3-o x o -
3a ,4,4a ,5,6,7,8,9a -oct a h yd r ob en z[f]isoin d ole-4-ca r b oxy-
lic Acid (6d ) (Figure 1, Condition B). Diene 3d (202 mg, 0.456
mmol) was treated with maleic anhydride (134 mg, 1.368
mmol) in CH2Cl2 as for 6c. The crude product was chromato-
graphed (140:100:12.5 hexane:CH2Cl2:AcOH) and 6d obtained
as a solid in 62% yield: mp 131-135 °C; [R]25D ) -60.4 (c 3.86,
CH2Cl2); 1H NMR δ 13.0 (bs, 1 H), 5.35, 5.29 (AB, J ) 12 Hz,
2 H, Bn), 5.19 (bs, 1 H, H9), 3.92 (m, 2 H, H10), 3.77 (m, 1 H,
H1), 3.73 (m, 1 H, H3a), 3.10 (dd, J ) 4, 6 Hz, 1 H, H4), 2.96
(m, 1 H, H9a), 2.46 (m, 1 H, H4a), 2.20 (m, 1 H, H8eq), 2.0 (m,
2 H, H5eq, H8ax), 1.8 (bs, 2 H, H6eq, H7eq), 1.57 (sep, J ) 7 Hz,
1 H, CH(CH3)2), 1.35-1.05 (m, 3 H, H7ax, H5ax, H6ax), 0.87 (d,
J ) 7 Hz, 6 H, CH(CH3)2), 0.82 (s, 6 H, SiC(CH3)2), 0.06 (s, 6
H, Si(CH3)2); 13C NMR δ 178.6, 174.3, 151.0, 145.5 (C8a), 134.8,
128.6, 128.6, 128.1 (Ar), 117.4 (C9), 68.6 (Bn), 64.9 (C10), 63.2
(C1), 45.1 (C4), 40.9 (C3a), 39.1 (C4a), 37.0 (C8), 36.5 (C9a),
34.0 (tertiary), 31.6 (C5), 28.8 (C7), 26.8 (C6), 25.0 (quater-
nary), 20.1, 20.1, 18.3, 18.3 (CH3), -3.8 (Si(CH3)2). Anal.
Calcd for C30H43NSiO6: C, 66.51; H, 8.00; N, 2.59. Found: C,
66.16; H, 8.27; N, 2.49.
(1S,3a S,4R,4a R,9a S)-1-(Acet oxym et h yl)-2-N-(ter t-b u -
toxyca r bon yl)-3-oxo-3a ,4,4a ,5,6,7,8,9a -octa h yd r oben z[f]-
isoin d ole-4-ca r boxylic Acid (6f). A solution of diene 3f (116
mg, 0.375 mmol) and maleic anhydride (110 mg, 1.122 mmol)
in CDCl3 (0.40 mL) was stirred at rt. When the reaction
mixture was analyzed by 1H NMR after 4 h, the ratio of the
resonances at δ ) 4.45 ppm (6f) and at δ ) 4.88 ppm (7f) was
3.6:1.0 in favor of 6f. Analysis of the reaction mixture after
24 h revealed that this ratio had not changed. The residue
was chromatographed (47:47:6 CH2Cl2:hexane:AcOH) to fur-
nish 6f as a solid (18%): mp 144-146 °C; [R]25 ) -55.6° (c
D
0.65,CH2Cl2); IR (Nujol) 3200-2400, 1775, 1745, 1700 cm-1
;
1H NMR δ 14.1 (bs, 1 H), 5.21 (bs, 1 H), 4.46 (dd, J ) 5, 12
Hz, 1 H), 4.23 (m, 1 H), 4.08 (m, 1 H), 3.53 (dd, J ) 4, 6 Hz, 1
H), 3.15 (dd, J ) 4, 6 Hz, 1 H), 2.89 (m, 1 H), 2.50 (m, 1 H),
2.27 (m, 1 H), 2.10 (s, 3 H, C(O)CH3), 2.0 (m, 2 H, H5eq, H8ax),
1.8 (bs, 2 H, H6eq, H7eq), 1.55 (s, 9 H, C(CH3)), 1.4 (m, 1 H,
H7ax), 1.25 (m, 2 H, H5ax, H6ax); 13C NMR δ 178.0, 173.9, 170.4,
149.0, 146.0 (C8a), 116.7 (C9), 84.9 (C(CH3)), 68.1 (Bn), 63.8
(C10), 62.1 (C1), 45.2 (C10), 40.6 (C3a), 39.0 (C4a), 36.5 (C8),
36.3 (C9a), 31.6 (C5), 28.9 (C7), 27.9 (C(CH3), 26.7 (C6), 20.4
(C(O)CH3). Anal. Calcd for C21H29NO7: C, 61.90; H, 7.17; N
3.44. Found: C, 61.59; H, 6.96; N, 3.36.
(1S,3a R,4S,4a S,9a R)-1-(Acetoxym eth yl)-2-N-(ben zoxy-
ca r bon yl)-3-oxo-3a ,4,4a ,5,6,7,8,9a -octa h yd r oben z[f]isoin -
d ole-4-ca r boxylic Acid (7a ) (Figure 1, Condition A).
A
mixture of diene 3a (500 mg, 1.456 mmol) and maleic anhy-
dride (428 mg, 4.368 mmol) in 1,2-dichlorobenzene (50 mL)
was refluxed for 1 h. An aliquot (5 mL) of the reaction mixture
was withdrawn, the solvent distilled (7 Torr), and the residue
analyzed by 1H and 13C NMR. No starting diene 7a was
detected, and resonances due to trans-isoindolone 6a and cis-
isoindolone 7a dominated each spectrum. The ratio of the
(Figure 1, Condition C). Diene 3d (212 mg, 0.479 mmol)
was treated with maleic anhydride (141 mg, 1.436 mmol) as

Diels-Alder Reactions of Chiral 1,3-Dienes
J . Org. Chem., Vol. 61, No. 24, 1996 8431
resonances at δ ) 3.57 ppm (6a ) and at δ ) 3.39 ppm (7a )
was 2.6:1.0. The workup was the same as for 6a . The crude
products were purified by HPLC, and a partially purified
sample of 7a was isolated: 1H NMR (incomplete) δ 13.3 (bs, 1
H), 7.42-7.33 (m, 5 H), 5.29 (dd, 2 H, Bn), 5.21 (bs, 1 H, H9),
4.87 (dd, J ) 4, 11 Hz, 1 H, H10), 4.33 (ddd, J ) 4, 7, 9 Hz, 1
H, H1), 4.19 (dd, J ) 9, 11 Hz, 1 H, H10), 3.37 (m, 1 H, CH3a),
2.49 (m, 1 H, H4a), 2.24 (m, 1 H, H8eq); 13C NMR (incomplete)
δ 112.0 (C9), 67.7 (Bn), 60.6 (C10), 56.8 (C1), 43.6 (C10), 41.1
(C3a), 38.0 (C4a), 36.6 (C8), 33.9 (C9a), 31.0 (C5), 28.6 (C7),
26.5 (C6), 20.7. From the main fraction pure 6a was isolated
in 30% yield.
(1S,3a R,4S,4a S,9a R)-1-[(2′,2′-Dim eth yleth oxy)m eth yl]-
2-N-(ben zoxyca r bon yl)-3-oxo-3a ,4,4a , 5,6,7,8,9a -octa h y-
d r oben z[f]isoin d ol-4-ca r boxylic Acid (7c) (Figure 1, Con-
dition A). Diene 3c (141 mg, 0.366 mmol) was treated with
maleic anhydride (107.8 mg, 1.099 mmol) in refluxing 1,2-
dichlorobenzene (12 mL) for 1 h. Chromatography (120:120:
12.5 hexane:CH2Cl2:AcOH) of the crude product afforded 7c
as an oil in 9% yield: 1H NMR δ 11.9 (bs, 1 H), 7.41 (m, 5 H),
5.31, 5.27 (AB, J ) 12 Hz, 2 H, Bn), 5.12 (bs, 1 H), 4.76 (dd, J
) 4,11 Hz, 1 H, H10), 4.39-4.26 (m, 2 H, H10, H1), 3.30 (dd,
J 3a,9a)7 Hz, J 3,4)4 Hz, 1 H, H3a), 3.19 (dd, J 4,4a)7 Hz, J 4,3a)4
Hz, 1 H, H4), 3.08 (m, 1 H, H9a), 2.53 (m, 1 H, H4a), 2.24 (m,
1 H, H8eq), 2.0 (m, 2 H, H5eq, H8ax), 1.8 (bs, 2 H, H6eq, H7eq),
1.4-1.1 (m, 3 H, H7ax, H6ax, H5ax), 1.16 (s, 9 H, CH3); 13C NMR
δ 178.6, 173.2, 151.3, 147.4 (C8a), 135.4, 128.7, 128.5, 128.5
(Ar), 111.8 (C9), 68.9 (Bn), 60.8 (C10), 57.8 (C1), 43.3 (C4),
41.4 (C3a), 38.9 (C4a), 38.9 (quaternary), 37.2 (C8), 35.0 (C9a),
31.4 (C5), 28.9 (C7), 27.1 (C6), 27.0 (CH3). Anal. Calcd for
C27H33NO7: C, 67.06; H, 6.88; N, 2.90. Found: C, 66.89; H,
6.92; N, 2.77. From the main fraction pure 6c was isolated in
36% yield.
(1S,3a R,4S,4a S,9a R)-1-[[Dim eth yl(1′,1′,2′-tr im eth ylp r o-
p y l)s i lo x y ]m e t h y l]-2-N -(b e n z o x y c a r b o n y l)-3-o x o -
3a ,4,4a ,5,6,7,8,9a -oct a h yd r ob en z[f]isoin d ole-4-ca r b oxy-
lic Acid (7d ) (Figure 1, Condition A). Diene 3d (168 mg, 0.379
mmol) was treated with maleic anhydride (112 mg, 1.138
mmol) as described for 7c. The crude product was chromato-
graphed (140:100:12.5 hexane:CH2Cl2:AcOH) and 7d obtained
as an oil (11%): 1H NMR δ 14.3 (bs, 1 H), 7.35 (m, 5 H), 5.46
(bs, 1 H, H9), 5.25 (dd, 2 H, Bn), 4.29 (dd, J ) 4, 10 Hz, 1 H,
H10), 4.08 (ddd, J ) 4, 6, 10 Hz, 1 H, H1), 3.64 (t, J ) 10 Hz,
1 H, H10), 3.25 (dd, J 3a,9a)7 Hz, J 3,4)4 Hz, 1 H, H3a), 3.14
(dd, J 4,4a)7 Hz, J 4,3a)4 Hz, 1 H, H4), 3.05 (m, 1 H, H9a), 2.47
(m, 1 H, H4a), 2.19 (m, 1 H, H8eq), 2.0 (m, 2 H, H5eq, H8ax),
1.8 (bs, 2 H, H6eq, H7eq), 1.57 (sep, J ) 7 Hz, 1 H, CH(CH3)2),
1.4-1.1 (m, 3 H, H7ax, H6ax, H5ax), 0.87 (d, J ) 7 Hz, 6 H,
CH(CH3)2), 0.83 (s, 6 H, SiC(CH3)2), 0.06 (s, 6 H, Si(CH3)2);
13C NMR δ 178.5, 173.2, 151.0, 146.3 (C8a), 134.6, 128.7, 128.5,
128.4 (Ar), 113.0 (C9), 68.9 (Bn), 60.5 (C10), 59.5 (C1), 44.4
(C4), 41.5 (C3a), 39.1 (C4a), 37.1 (C8), 34.6 (C9a), 34.0, 31.5
(C5), 29.0 (C7), 27.0 (C6), 25.0, 20.1, 20.1, 18.3, 18.3 (CH3),
-3.7 (Si(CH3)2). From the main fraction pure 6e was isolated
in a yield of 34%.
132.9, 132.8, 130.1, 130.0, 128.6, 128.4, 128.4, 127.9, 112.7
(C9), 68.9 (Bn), 60.4 (C10), 60.7 (C1), 45.0 (C4), 41.6 (C3a),
39.2 (C4a), 37.0 (C8), 34.7 (C9a), 31.6 (C5), 29.8, 28.8 (C7),
26.9 (C6), 26.9 (CH3). Anal. Calcd for C38H43NSiO6: C, 71.56;
H, 6.80; N, 2.20. Found: C, 71.34; H, 6.81; N, 2.17. From
the main fraction pure 6e was isolated in 39% yield.
Met h yl (1S,3a S,4R,4a R,9a S)-1-(Acet oxym et h yl)-2-N-
(ben zoxyca r bon yl)-3-oxo-3a ,4,4a ,5,6,7,8,9a -octa n on a h y-
d r ob en z[f]isoin d ole-4-ca r b oxyla t e (6g) a n d Met h yl
(1S,3a R,4S,4a S,9a R)-1-(Acet oxym et h yl)-2-N-(b en zoxy-
ca r bon yl)-3-oxo-3a ,4,4a ,5,6,7,8,9a -octa h yd r oben z[f]isoin -
d ole-4-ca r boxyla t e (7g). A solution of diene 3a (500 mg,
1.456 mmol) and maleic anhydride (428 mg, 4.368 mg) was
refluxed in 1,2-dichlorobenzene (50 mL) as for 7a . The crude
acidic extracts (826 mg) were dissolved in ether (50 mL) and
cooled to -10 °C in a conical flask. To this stirred solution
was slowly added an ethereal solution of CH2N2 until the
yellow color just persisted. A few drops of AcOH were added
to discolor the solution. The solvent was evaporated and the
residue dissolved in a minimum amount of MeOH and filtered
through a 3 cm thick pad of reversed-phase silica using MeOH
(100 mL) as the eluting solvent. After the solvent was
evaporated from the filtrate the residue was dissolved in a
minimum amount of EtOAc and filtered through a 4 cm thick
pad of H60 silica (Merck) using EtOAc (150 mL) as the eluting
solvent. The solvent was evaporated from the filtrate to give
a brown oil (491 mg, 74% with regard to diene 3a ). 1H and
13C NMR spectroscopic analysis of the oil revealed a ratio of
2.7:1.0 for 6g and 7g. An aliquot of the crude product was
analyzed by HPLC. Two peaks were detected with retention
times of t ) 10.9 min (7g) and t ) 12.1 min (6g) in a ratio of
1.0:2.7, respectively. The crude product was dissolved in a
minimum amount of hot MeOH (2 mL), and the solution was
allowed to stand at rt for 1 h. The mother liquors were
removed and the remaining crystals washed with ice-cold
MeOH (1 mL) and dried to yield 54 mg of 7g (8% with respect
to diene 3a ): mp 202-203 °C; [R]25 ) 120° (c 1.37, CDCl3);
D
IR (Nujol) 1760, 1740 cm-1; H NMR δ 12.2 (bs, 1 H), 7.44-
1
7.31 (m, 5 H), 5.31, 5.26 (AB, J ) 12 Hz, 2 H, Bn), 5.21 (bs, 1
H, H9), 4.86 (dd, J ) 4,10 Hz, 1 H, H10), 4.30 (ddd, J ) 4, 7,
9 Hz, 1 H, H1), 4.19 (dd, J ) 9, 10 Hz, 1 H, H10), 3.78 (s, 3 H,
OCH3), 3.46 (dd, J ) 4, 8 Hz, 1 H, H3a), 3.12 (m, 1 H, H9a),
2.92 (dd, J ) 4, 6 Hz, 1 H, H4), 2.46 (m, 1 H, H4a), 2.25 (m, 1
H, H8eq), 2.05 (s, 3 H, C(O)CH3), 2.0 (m, 2 H, H5eq, H8ax), 1.8
(m, 2 H, H6eq, H7eq), 1.4-1.1 (m, 3 H, H7ax, H6ax, H5ax); 13C
NMR δ 173.6, 171.9, 170.3 (CO), 152.0 (NC(O)O), 147.6
(CHdC), 135.1 (quaternary ar), 128.6, 128.4, 128.4, 112.2
(CdCH), 68.5 (Bn), 61.4 (CH2O), 57.2 (CHN), 51.9 (OCH3), 41.3
(CHCO2H), 41.2 (CHC(O)N), 38.8 (CdCCH), 37.4 (CdCCH2),
34.3 (CdCHCH), 31.4, 29.4, 27.2, 20.8 (C(O)CH3). Anal. Calcd
for C25H29NO7: C, 65.92; H, 6.42; N, 3.07. Found: C, 65.75;
H, 6.42; N, 2.95.
Diastereomeric 6g was obtained (21% with respect to diene
3a ) as a colorless oil by HPLC of the mother liquor: [R]25
)
D
-45.2° (c 1.63, CH2Cl2); IR (neat) 1795,1745-1730, 1700 cm-1
;
(1S,3a R ,4S,4a S,9a R )-1-[(Dip h e n yl-t er t -b u t ylsiloxy)-
m eth yl]-2-N-(ben zoxyca r bon yl)-3-oxo-3a ,4,4a ,5, 6,7,8,9a -
octa h yd r oben z[f]isoin d ole-4-ca r boxylic Acid (7e). (Fig-
ure 1, Condition A). Diene 3e (231 mg, 0.406 mmol) was
treated with maleic anhydride (118 mg, 1.208 mmol) as
described for 7c. The crude product was chromatographed
(140:100:12.5 hexane:CH2Cl2:AcOH) and 7e obtained as an oil
1H NMR δ 7.38-7.30 (m, 5 H), 5.29, 5.26 (AB, J ) 12 Hz, 2 H,
Bn), 5.19 (bs, 1 H, H9), 4.39 (dd, J ) 5, 12 Hz, 1 H, H10), 4.21
(dd, J ) 4, 12 Hz, 1 H, H10), 4.11 (m, 1 H, H1), 3.77 (s, 3 H,
OCH3), 3.69 (dd, J ) 4, 8 Hz, 1 H, H3a), 2.89 (m, 1 H, H4,),
2.88 (m, 1 H, H9a), 2.44 (m, 1 H, H4a), 2.25 (m, 2 H, H8eq
,
H5eq), 2.05 (s, 3 H, C(O)CH3), 1.97 (m, 1 H, H8ax), 1.75 (m, 2
H, H6eq, H7eq), 1.4 (m, 1 H, H7ax), 1.21 (m, 1 H, H6ax), 1.05 (m,
1 H, H5ax); 13C NMR δ 173.1, 171.7, 170.4, 151.5, 146.4 (C8a),
135.2, 128.8, 128.5, 127.9 (Ar), 116.5 (C9), 68.2 (Bn), 63.8 (C10),
60.8 (C1), 51.7 (OCH3), 42.1 (C4), 40.2 (C3a), 38.4 (C4a), 36.7
(C8), 36.0 (C9a), 31.5 (C7), 29.3 (C7), 27.1 (C6), 20.7 (C(O)-
CH3). Anal. Calcd for C25H29NO7: C, 65.92; H, 6.42; N 3.07.
Found: C, 65.68; H, 6.62; N, 2.98.
(11%): [R]25 ) -3.26° (c 0.46,CH2Cl2); IR (neat) 3300-2500,
D
1795, 1760, 1730 cm-1; H NMR δ 13.8 (bs, 1 H), 7.65-7.25
1
(m, 15 H), 5.51 (bs, 1 H, H9), 5.10 (dd, 2 H, Bn), 4.45 (dd, J )
5, 10 Hz, 1 H, H10), 4.11 (m, 1 H, H1), 3.72 (t, J ) 10 Hz, 1 H,
H10), 3.28-3.18 (m, 2 H, H3a, H4, H9a), 2.52 (m, 1 H, H4a),
2.15 (m, 1 H, H8eq), 2.0 (m, 2 H, H5eq, H8ax), 1.8 (bs, 2 H, H6eq,
H7eq), 1.4-1.1 (m, 3 H, H7ax, H6ax, H5ax), 1.07 (s, 9 H, CH3);
13C NMR δ 178.9, 173.4, 150.9, 146.7 (C8a), 135.5, 135.5, 134.3,
J O9609743