Tandem [4 + 2]/[3 + 2] cycloadditions of nitroalkenes. 11. The synthesis of (+)-crotanecine

Article abstract of DOI:10.1021/ja963084d

(+)-Crotanecine (1) is the necine base component of a number of pyrrolizidine alkaloids. This necine subunit is an amino triol bearing a primary allylic alcohol characterized by an all-cis relationship of its stereocenters. The synthesis of (+)-crotanecine has been accomplished in 10 steps and 10.2% overall yield. The key step in the asymmetric synthesis is a Lewis acid-promoted, tandem inter[4 + 2]/intra[3 + 2] cycloaddition between a (fumaroyloxy)nitroalkene 14 and chiral β-silylvinyl ether (-)-26. This synthesis serves to illustrate the synthetic versatility of the tandem cycloaddition to incorporate additional functionality.

Full text of DOI:10.1021/ja963084d

J. Am. Chem. Soc. 1997, 119, 125-137
125
Tandem [4 + 2]/[3 + 2] Cycloadditions of Nitroalkenes. 11.
The Synthesis of (+)-Crotanecine
Scott E. Denmark* and Atli Thorarensen
Contribution from Roger Adams Laboratory, Department of Chemistry, UniVersity of Illinois,
Urbana, Illinois 61801
ReceiVed September 3, 1996X
Abstract: (+)-Crotanecine (1) is the necine base component of a number of pyrrolizidine alkaloids. This necine
subunit is an amino triol bearing a primary allylic alcohol characterized by an all-cis relationship of its stereocenters.
The synthesis of (+)-crotanecine has been accomplished in 10 steps and 10.2% overall yield. The key step in the
asymmetric synthesis is a Lewis acid-promoted, tandem inter[4 + 2]/intra[3 + 2] cycloaddition between a
(fumaroyloxy)nitroalkene 14 and chiral â-silylvinyl ether (-)-26. This synthesis serves to illustrate the synthetic
versatility of the tandem cycloaddition to incorporate additional functionality.
Introduction and Background
Synthetic efficiency is currently one of the most important
challenges in organic chemistry.1 In recent years the strategic
use of tandem reactions has been well recognized as a powerful
method for increasing molecular complexity and thereby
Figure 1. Necine bases that have been prepared with the tandem [4
+ 2]/[3 + 2] cycloadditions.
synthetic utility.2,3 The tandem [4 + 2]/[3 + 2] cycloadditions
of nitroalkenes have been extensively developed in these
laboratories for the construction of various polycyclic nitrogen-
containing compounds in enantiomerically enriched form.4
Indeed, this approach has been successfully employed in the
synthesis of the two pyrrolizidine alkaloids (-)-hastanecine5
and (-)-rosmarinecine.6
Australia six new alkaloids were isolated, all containing (+)-
crotanecine as the base (Figure 2).10 The structure of anacrotine
was unambiguously proven by X-ray crystallographic analysis
in 1984, confirming the relative and absolute stereochemistry
and the size of the macrolactone.11 In addition, an X-ray crystal
structure of (+)-crotanecine itself has been determined.12 Thus,
the structure of (+)-crotanecine is firmly secured. Until
recently,13 the only 1H NMR data available for (+)-crotanecine
As a part of a continuing program in methodology-driven
alkaloid synthesis, we recognized a challenge provided by such
molecules as (+)-crotanecine which bear an additional hydroxyl
substituent at C(2). The construction of such a molecule by
the tandem sequence would require the ability to install this
functionality in the correct configuration by the use of a
functionalized chiral dienophile. The evolution of a general
approach to these compounds and the successful synthesis of
(+)-crotanecine are described below.
(+)-Crotanecine is found conjugated to a variety of necic
acids in many pyrrolizidine alkaloids found in plants of the
Crotalaria species. For example, anacrotine (4) isolated from
the seeds of C. anagyroids obtained from Sri Lanka7,8 or from
C. incana shrub grown in South Africa9 afforded senecic acid
and a new amino triol named crotanecine. Upon reexamination
of leaves and twigs gathered from C. agatiflora grown in
1
were from the original 60 MHz H NMR spectrum.7
(+)-Crotanecine bears a double bond at the C(6)-C(7)
position; therefore, all the alkaloids 4-11 that contain crotane-
cine should be hepatotoxic, the most common biological effect
of pyrrolizidine alkaloids. However, this has only been
demonstrated for anacrotine.14 In addition to hepatotoxicity,
anacrotine displays pneumotoxicity.
There have been only two previous syntheses of (+)-
crotanecine15,16 along with one formal synthesis.17 They all
(8) (a) Culvenor, C. C. J.; Smith, L. W.; Willing, R. I. Chem. Commun.
1970, 65. Anacrotine has in addition been isolated from C. laburnifolia:
(b) Sawhney, R. S.; Girotra, R. N.; Atal, C. K.; Culvenor, C. C. J.; Smith,
L. W. Indian J. Chem. 1967, 5, 655. (c) Crout, D. H. G. J. Chem. Soc.,
Perkin Trans. 1 1972, 1602.
(9) This report contains the only reported physical data of analytically
pure crotanecine: Mattocks, A. R. J. Chem. Soc. C 1968, 235.
(10) Culvenor, C. C. J.; Smith, L. W. Anal. Quim. 1972, 68, 883.
(11) Mackay, M. F.; Sadek, M.; Culvenor, C. C. J. Acta Crystallogr. C
1984, 40, 1073.
(12) Richardson, J. F.; Culvenor, C. C. J. Acta Crystallogr. C 1985, 41,
1475.
(13) Ro¨der, E.; Liu, K.; Bourauel, T. Fresenius’ J. Anal. Chem. 1992,
342, 719.
(14) (a) Mattocks, A. R.; Driver, H. E. Chem.-Biol. Interact. 1987, 63,
91. (b) Mattocks, A. R.; Bird, I. Chem.-Biol. Interact. 1983, 43, 209. (c)
Mattocks, A. R.; Jukes, R. Nat. Toxins 1992, 1, 89.
(15) Yadav, V. K.; Ru¨eger, H.; Benn, M. Heterocycles 1984, 22, 2735.
(16) Bennett, R. B. III: Cha, J. K. Tetrahedron Lett. 1990, 31, 5437.
(17) Buchanan, J. G.; Jigajinni, V. B.; Singh, G.; Wightman, R. H. J.
Chem. Soc., Perkin Trans. 1 1987, 2377.
X Abstract published in AdVance ACS Abstracts, December 15, 1996.
(1) Hall, N. Science 1994, 266, 32.
(2) (a) Tietze, L. F.; Beifuss, U. Angew. Chem., Int. Ed. Engl. 1993, 32,
131. (b) Tietze, L. F.; Bachmann, J.; Wichmann, J.; Burkhardt, O. Synthesis
1994, 1185. (c) Bunce, R. A. Tetrahedron 1995, 51, 13103. (d) Tietze, L.
F. Chem. ReV. 1996, 96, 115. (e) Winkler, J. D. Chem. ReV. 1996, 96,
167. (f) Ryu, I.; Sonoda, N.; Curran, D. P. Chem. ReV. 1996, 96, 177. (g)
Parsons, P. J.; Penkett, C. S.; Shell, A. J. Chem. ReV. 1996, 96, 195. (h)
Wang, K. K. Chem. ReV. 1996, 96, 207. (i) Padwa, A.; Weingarten, M. D.
Chem. ReV. 1996, 96, 223.
(3) (a) Ho, T.-L. Tandem Organic Reactions; Wiley: New York, 1992.
(b) Thebtaranonth, C.; Thebtaranonth, Y. Cyclization Reactions; CRC
Press: Boca Raton, FL, 1994.
(4) Denmark, S. E.; Thorarensen, A. Chem. ReV. 1996, 96, 137.
(5) Denmark, S. E.; Thorarensen, A. J. Org. Chem. 1994, 59, 5672.
(6) Denmark, S. E.; Thorarensen, A.; Middleton, D. S. J. Org. Chem.
1995, 60, 3574.
(7) Atal, C. K.; Kapur, K. K.; Culvenor, C. C. J.; Smith, L. W.
Tetrahedron Lett. 1966, 537.
S0002-7863(96)03084-3 CCC: $14.00 © 1997 American Chemical Society

126 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Denmark and Thorarensen
Figure 3. Examples of different necines organized by the classifications
of the relationship between C(1), C(7), and C(7a).
Scheme 1
Figure 2. Alkaloids that contain (+)-crotanecine (1) as the necine base.
share in common the use of the chiral pool starting material
that incorporates one or more of the stereocenters found in (+)-
crotanecine.
Synthesis Design18
The tandem [4 + 2]/[3 + 2] cycloaddition of nitroalkenes is
an extremely flexible method for the synthesis of necines. All
of the stereochemical attributes are subject to a high level of
predictability and control. Further, the installation of functional
groups at key stereocenters can be achieved by appropriate
modification of diene and dienophile. Finally, judicious choice
of chiral auxiliary and Lewis acid sets the absolute configuration
of the molecule as a whole.
The power of this strategy becomes apparent when classifying
all the necines on the basis of the relationship among the
stereogenic centers at C(1), C(7), and C(7a) (Figure 3). There
are 21 structurally unique 7-hydroxymethyl-substituted necines.
Ten of those have the all-cis relationship exemplified by (-)-
rosmarinecine and could arise from a tandem inter[4 + 2]/intra[3
+ 2] process. Another seven necines have the all-trans
relationship as in (-)-hastanecine and could arise from a tandem
inter[4 + 2]/inter[3 + 2] process. In addition there are two
members which contain a cis/trans relationship, leaving only
two necines conceptually not accessible with this method.
(+)-Crotanecine belongs to the family of all-cis-substituted
necines. Thus, the retrosynthetic analysis of (+)-crotanecine
follows the same logic as was employed for (-)-rosmarinecine;6
however, it represented a significant synthetic challenge due to
the higher level of functionality, namely the C(6)-C(7) olefin
and the C(2) hydroxyl group (Scheme 1). (+)-Crotanecine was
envisioned to arise from the R-hydroxy lactam 16, where X
would be an appropriate hydroxyl synthon. The hydroxyl at
C(1) then becomes the handle necessary to install the olefin at
C(6)-C(7) in (+)-crotanecine. The R-hydroxy lactam 16 would
derive from a nitroso acetal 15 after reductive unmasking. The
nitroso acetal 15 in turn arises from an inter[4 + 2]/intra[3 +
2] cycloaddition of the nitroalkene 14 and an appropriately
substituted propenyl ether. The ideal propenyl ether would be
the 2-acetoxyvinyl ether 13, which was utilized in the tandem
intra[4 + 2]/intra[3 + 2] cycloaddition approach to the synthesis
of (+)-pretazettine and in the synthesis of 4-hydroxy-substituted
pyrrolidines.19 Since the 2-acetoxyvinyl ether 13 is only easily
accessible in the cis configuration, it demands that the cyclo-
addition with nitroalkene 14 must occur through an endo
approach to install the correct relative configuration at C(4a)
and C(5) in the resulting nitroso acetal (C(1) and C(2) in (+)-
crotanecine). From the outset there were three questions which
had to be addressed: (1) would the nitroalkene 14 allow an
endo approach of the dienophile in the cycloaddition with 13,
(2) could any Lewis acid other than tin tetrachloride promote
the cycloaddition with 13 and thereby alter the inherent exo
preference for the approach of the dienophile 13, and (3) if
unsuccessful, could other dienophiles bearing hydroxyl group
equivalents be devised and enlisted into service?
(18) For recent reviews on the synthesis and the biological properties of
pyrrolizidines see: (a) Natural Occuring Pyrrolizidine Alkaloids; Rizk,
A.-F. M., Ed.; CRC: Boston, MA, 1991. (b) Pyrrolizidine Alkaloids; World
Health Organization: Geneva, 1988. (c) Mattocks, A. R. Chemistry and
Toxicology of Pyrrolizidine Alkaloids; Academic Press: New York, 1986.
(d) Robins, D. J. Nat. Prod. Rep. 1995, 12, 413. (e) Ikeda, M.; Sato, T.;
Ishibashi, H. Heterocycles 1988, 27, 1465. (f) Dai, W.-M.; Nagao, Y.
Heterocycles 1990, 30, 1231. (g) Wro´bel, J. T. The Alkaloids: Brossi, A.,
Ed.; Academic Press: New York, 1985; Vol. 26, Chapter 7.
Results
First-Generation Approach with 2-Acetoxyvinyl Ether 13.
The nitroalkene 17 was used as a model substrate in cycload-
(19) Denmark, S. E.; Schnute, M. E. J. Org. Chem. 1994, 59, 4576.

Synthesis of (+)-Crotanecine
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 127
Table 1. Survey of Lewis Acids Promoters in Cycloadditions with
13
Scheme 2
entry Lewis acid temperature, °C diastereomer ratio yield, %
1
2
3
4
SnCl4
Ti(Oi-Pr)2Cl2
MAPh
-78
-40
-20
-78
13/1/1/1
1/1/1
34
32
0
trace
TiCl4
1/1
ditions with 13 due to its simpler preparation. These reactions
were promoted by selected Lewis acids used previously in the
[4 + 2] cycloaddition with nitroalkenes as heterodienes (Table
1). The initial results were, to some extent, encouraging since
a nitroso acetal 18 was isolated from the cycloadditions
promoted by Ti(Oi-Pr)2Cl2. Unfortunately, both the yield and
the selectivity were poor (entry 2). MAPh was found not to
promote this reaction, while tin tetrachloride afforded the
cycloadduct with moderate selectivity. In the cycloaddition,
all four possible diastereomers were obtained with one dominat-
ing in a 13/1/1/1 ratio as determined by 1H NMR analysis. These
nitroso acetals were also isolated in poor yield, and changing
the amount or addition order of Lewis acid afforded no
improvements. By thin layer chromatographic analysis, the
vinyl ether was consumed faster by reaction with tin tetrachlo-
ride than by cycloaddition. This was confirmed by isolation
of a side product derived from chloride addition to the enol
ether which was noted previously.20 For these reasons, and the
ambiguity about the relative configuration at C(4a)-C(5), this
approach was abandoned and attention redirected to the use of
other propenyl ethers bearing a hydroxyl surrogate.
Second-Generation Approach with a 2-Silylvinyl Ether.
The need for a hydroxyl surrogate directed attention to the use
of substituted silanes as hydroxyl synthons.21-23 The silyl group
has been used on numerous occasions in organic synthesis to
modulate the reactivity of functional groups.24 Therefore the
use of a vinylsilane as the dienophile in the tandem [4 + 2]/[3
+ 2] cycloaddition was a tempting alternative to the 2-acetoxy-
vinyl ether 13. Indeed, vinylsilanes have previously been used
as enol surrogates in the Diels-Alder reaction.25
in great detail before the synthesis was undertaken (Scheme
2). (+)-Crotanecine was envisioned to arise from the R-hydroxy
lactam 23 where deprotection of the methyl acetal followed by
elimination of the mesylate and exhaustive reduction would
afford the natural product. The R-hydroxy lactam 23 would
arise from 22 by a series of transformations beginning with the
differentiation of the two hydroxyl groups. Transformation of
the lactol in 22 to a methyl acetal, followed by mesylation and
unmasking the silyl group by the Tamao-Fleming protocol21,22
should produce 23. The R-hydroxy lactam 22 would arise from
nitroso acetal 21 after the two-step unmasking procedure. The
nitroso acetal 21 itself is the result of an inter[4 + 2]/intra[3 +
2] cycloaddition of the nitroalkene 14 and a 2-silylvinyl ether.
In this instance the isomer of 2-silylvinyl ether 19 was selected,
since the nitroalkene 14 had been successfully utilized in the
tandem [4 + 2]/[3 + 2] cycloadditions only with MAPh as the
Lewis acid promoter. Since MAPh promotes an exo-mode [4
+ 2] cycloaddition, a trans-2-silylvinyl ether is required to
ensure the correct cis relationship at C(4a)-C(5) in the product
nitroso acetal (C(1)-C(2) in (+)-crotanecine).
Since both methyl and phenyl propenyl ethers had been
employed with great success in the [4 + 2] cycloaddition of
nitroalkenes, silyl substitution was not expected to dramatically
alter the reactivity of the vinyl ether.26 Therefore, the approach
to (+)-crotanecine using a 2-silylvinyl ether could be outlined
Synthesis of the Desired (E)-2-Silylvinyl Ether. The
required (E)-2-silylvinyl ether 26 was to be prepared by the
method of Green27 from the 2-silylpropynyl ether 25 followed
by an LiAlH4 reduction.28 Attempts to prepare the desired
(25) (a) Taber, D. F.; Bhamidipati, R. S.; Yet, L. J. Org. Chem. 1995,
60, 5537. (b) Kolb, H. C.; Ley, S. V.; Slawin, A. M. Z.; Williams, D. J.
J. Chem. Soc., Perkin Trans. 1 1992, 2735. (c) Sieburth, S. M.; Fensterbank,
L. J. Org. Chem. 1992, 57, 5279. (d) Stork, G.; Chan, T. Y.; Breault, G.
A. J. Am. Chem. Soc. 1992, 114, 7578. (e) Tamao, K.; Kobayashi, K.; Ito,
Y. J. Am. Chem. Soc. 1989, 111, 6478. (f) Chen, R.-M.; Weng, W.-W.;
Luh, T.-Y. J. Org. Chem. 1995, 60, 3272. (g) Shea, K. J.; Zandi, K. S.;
Staab, A. J.; Carr, R. Tetrahedron Lett. 1990, 31, 5885. (h) Singleton, D.
A.; Redman, A. M. Tetrahedron Lett. 1994, 35, 509. (i) Koreeda, M.; Teng,
K.; Murata, T. Tetrahedron Lett. 1990, 31, 5997.
(20) Schnute, M. E. Personal communication.
(21) (a) Fleming, I.; Henning, R.; Plaut, H. J. Chem. Soc., Chem.
Commun. 1984, 29. (b) Chow, H.-F.; Fleming, I. Tetrahedron Lett. 1985,
29, 397. (c) Fleming, I.; Hill, J. H. M.; Parker, D.; Waterson, D. J. Chem.
Soc., Chem. Commun. 1985, 318.
(22) (a) Tamao, K.; Ishida, N. J. Organomet. Chem. 1984, 269, C37.
(b) Tamao, K.; Ishida, N.; Tanaka, T.; Kumada, M. Organometallics 1983,
2, 1694. (c) Tamao, K.; Kumada, M.; Maeda, K. Tetrahedron Lett. 1984,
25, 321.
(23) For reviews oxidation of C-Si bonds, see: (a) Colvin, E. W. In
ComprehensiVe Organic Synthesis: Ley, S. V., Ed.; Pergamon Press:
Oxford, U.K., 1991; Vol. 7, Chapter 4.3. (b) Fleming, I. Chemtracts: Org.
Chem. 1996, 9, 1.
(24) For recent reviews concerning the use of silicon building blocks in
organic synthesis, see: (a) White, J. M. Aust. J. Chem. 1995, 48, 1227. (b)
Hwu, J. R.; Patel, H. V. Synlett 1995, 989. (c) Luh, T.-Y.; Wong, K.-T.
Synthesis 1993, 349.
(26) Fleming, I. Chem. Soc. ReV. 1981, 10, 83.
(27) (a) Moyano, A.; Charbonnier, F.; Greene, A. E. J. Org. Chem. 1987,
52, 2919. (b) Charbonnier, F.; Moyano, A.; Greene, A. E. J. Org. Chem.
1987, 52, 2303.
(28) (a) Denmark, S. E.; Senanayake, C. B. W. J. Org. Chem. 1993, 58,
1853. (b) Denmark, S. E.; Schnute, M. E.; Senanayake, C. B. W. J. Org.
Chem. 1993, 58, 1859. (c) Denmark, S. E.; Marcin, L. R. J. Org. Chem.
1995, 60, 3221.

128 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Denmark and Thorarensen
Table 2. Temperature Effect in the Silylcupration of 24
Scheme 3
PhMe2SiLi,
equiv
temperature,
isomer
ratio, R/â
yield,
%
CuX
°C
2
2
2
1
2
CuBr‚DMS
CuI
CuCN
CuCN
CuCN
-78
-78
-78
0
1/1.9
1/2
1/2.2
1/2.3
0/1
92
81
69
23
85
0
acetylenes.38 The silylcuprate adds with excellent regio- and
stereoselectivity to afford the vinylsilanes through an exclusive
syn addition across the acetylene. Since there was no precedent
for the additions of silylcuprates to acetylenic ethers, a survey
of copper source and stoichiometry was performed (Table 2).
Lithiodimethylphenylsilane was prepared according to the
methods of Gilman and Fleming.38-40 All of the cuprates were
prepared by the addition of the silyllithium to the copper reagent
in tetrahydrofuran at 0 °C. After being stirred for 20 min, the
reaction mixture was cooled to the appropriate temperature. The
acetylenic ether 24 was then added as a THF solution, then was
stirred for 20 min (-78 °C) or 1.5 h (0 °C) and worked up to
afford the vinylsilane. Irrespective of the copper source, the
cuprations at -78 °C all afforded the vinylsilane in good yield
but in all instances as a mixture of R and â addition products
(28/26) in a ratio of 1/2 favoring the desired (E)-vinylsilane
26. All subsequent reactions were performed with copper
cyanide since it is much easier to handle than the other copper
salts. At 0 °C, the use of a lower order cuprate afforded no
improvements, while the higher order cyanocuprate gave the
desired (E)-vinylsilane 26 exclusively in 85% yield. This
reaction could be scaled up to afford multigram quantities of
the vinylsilane (E)-26. The vinylsilane was stable enough to
be purified on both basic alumina and silica gel. However, the
residual acid in deuterochloroform could isomerize the vinyl-
silane to a mixture of E and Z isomers. With the desired
vinylsilane in hand, attention was directed to establishing its
utility as a diene in the tandem cycloaddition sequence.
silylpropynyl ether in a one-pot reaction, from 2-phenylcyclo-
hexanol failed, and only phenylcyclohexenes were isolated.
Therefore the acetylenic ether 24 was prepared as described by
Porter.29 The acetylenic ether 24 was then deprotonated with
n-BuLi at -78 °C, and the resulting lithioalkyne was treated
with chlorodimethylphenylsilane to afford the silylated acetylene
25 in 85% yield (Scheme 3). Subjecting the silylpropenyl ether
25 to an LiAlH4 or Red-Al reduction afforded none of the
desired product.30,31 A survey of other aluminum- or chromium-
based reducing reagents capable of reducing propenyl com-
pounds to (E)-alkenes also met with no success.32-34 The only
reagent capable of reducing the propynyl compound was
DIBAL-H, which afforded exclusively the (Z)-2-silylvinyl ether
27 in 69% yield.35
An alternative, the addition of a silyl anion across the
acetylenic ether 24, was next examined. There are several
examples of the use of carbocupration of heteroatom (N or O)-
substituted alkynes wherein the preferred regiochemistry of
addition places the copper â to the heteroatom.36,37 Fleming
has demonstrated the synthetic utility of higher order silyl-
cyanocuprates in additions to nonfunctionalized monosubstituted
Optimization of the Tandem [4 + 2]/[3 + 2] with (E)-
(-)-26. The 2-silylvinyl ether (E)-(-)-26 was subjected to
cycloaddition with nitroalkene 14 in the presence of MAPh as
the Lewis acid promoter (Scheme 4). This vinyl ether was
found to be rather unreactive, affording only a poor yield of
the desired nitroso acetal at -78 °C (Table 3, entry 1).
Increasing the reaction temperature to 0 °C increased the yield
to 54%, but reactions conducted in methylene chloride at this
temperature became dark brown and, when quenched, no color
change nor gas evolution was observed, indicating the absence
of active Lewis acid. A change of solvent to toluene avoided
this problem, and the reactions were much cleaner. Further
adjustments of stoichiometry and temperature are summarized
in Table 3. The optimum reaction conditions for the cycload-
dition (entry 10) were found to be the use of 3 equiv of vinyl
ether (-)-26 and 5 equiv of MAPh in toluene at -14 °C. The
nitroso acetal (+)-29 was isolated in 73% yield in greater than
(29) Porter, N. A.; Dussault, P.; Breyer, R. A.; Kaplan, J.; Morelli, J.
Chem. Res. Toxicol. 1990, 3, 236.
(30) Sola´, L.; Castro, J.; Moyano, A.; Perica´s, M. A.; Riera, A.
Tetrahedron Lett. 1992, 33, 2863. For reductions of propynyl ethers to
E-propenyl ethers, see also ref 28.
(31) For aluminum-based reductions of propargyl compound to E-alkenes,
see: (a) Corey, E. J.; Kirst, H. A.; Katzenellenbogen, J. A. J. Am. Chem.
Soc. 1970, 92, 6314. (b) Zweifel, G.; Lewis, W.; On, H. P. J. Am. Chem.
Soc. 1979, 101, 5101. (c) Denmark, S. E.; Jones, T. K. J. Org. Chem.
1982, 47, 4595. (d) Marshall, J. A.; Shearer, B. G.; Crooks, S. L. J. Org.
Chem. 1987, 52, 1236. (e) Kang, M. J.; Jang, J.-S.; Lee, S.-G. Tetrahedron
Lett. 1995, 36, 8829. (f) Zweifel, G.; Steele, R. B. J. Am. Chem. Soc. 1967,
89, 5085.
(32) For chromium-based reducing agents, see: (a) Castro, C. E.;
Stephens, R. D. J. Am. Chem. Soc. 1964, 86, 4358. (b) Smith, A. B., III;
Levenberg, P. A.; Suits, J. Z. Synthesis 1986, 184. (c) Constantino, M. G.;
Donate, P. M.; Petragnani, N. J. Org. Chem. 1986, 51, 253.
(33) For reductions in NH3, see: (a) Schwarz, M.; Waters, R. M.
Synthesis 1972, 567. (b) Warthen, J. D., Jr.; Jacobson, M. Synthesis 1973,
616. (c) Boland, W.; Hansen, V.; Jaenicke, L. Synthesis 1979, 114.
(34) For a review on the reduction of triple-bonded groups, see: Hutchins,
R. O.; Hutchins, M. G. K. In The Chemistry of the Functional Groups,
Supplement C; Patai, S., Rappoport, Z., Eds; John Wiley: New York, 1983;
Chapter 14, p 571.
1
a 50/1 diastereomeric ratio of nitroso acetals as judged by H
NMR analysis. In addition, the chiral vinyl ether (-)-26 (65%)
(37) For recent reviews of cupration in organic synthesis, see: (a)
Lipshutz, B. H.; Sengupta, S. Org. React. 1992, 41, 135. (b) Lipshutz, B.
H. In Organometallics in Synthesis-A Manual; Schlosser, M., Ed.; John
Wiley: New York, 1994; Chapter 4, p 283.
(38) Fleming, I.; Newton, T. W.; Roessler, F. J. Chem. Soc., Perkin Trans.
1 1981, 2527.
(39) (a) George, M. V.; Peterson, D. J.; Gilman, H. J. Am. Chem. Soc.
1960, 82, 403. (b) Gilman, H.; Lichtenwalter, G. D. J. Am. Chem. Soc.
1958, 80, 608.
(35) For the use of DIBAL-H for the preparation of either (E)- or (Z)-
alkenes, see: (a) Zweifel, G.; On, H. P. Synthesis 1980, 803. (b) Eisch, J.
J.; Foxton, M. W. J. Org. Chem. 1971, 36, 3520. (c) Granitzer, W.; Stu¨tz,
A. Tetrahedron Lett. 1979, 3145.
(36) For reviews of carbocuparation, see: (a) Normant, J. F.; Alexakis,
A. Synthesis 1981, 841. (b) Casson, S.; Kocienski, P. Contemp. Org. Synth.
1995, 2, 19.
(40) For a review of silyl anions, see: Tamao, K.; Kawachi, A. In
AdVances in Organometallic Chemistry; Stone, F. G. A., West, R., Eds.;
Academic Press: New York, 1995; Vol. 38, Chapter 1, p 1.

Synthesis of (+)-Crotanecine
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 129
Scheme 4
Table 3. Optimization of Reaction Parameters in the [4 + 2]/[3 +
2] Cycloaddition
(-)-26,
equiv
MAPh,
equiv
temperature,
yield,
%
entry
°C
solvent
1a
2
3
4
5
6
7
8
9
3
3
3
3
3
3
1.5
3
3
3
3
3
1.2
6
8
6
6
6
-78
-40
0
CH2Cl2
CH2Cl2
toluene
toluene
toluene
toluene
toluene
CH2Cl2
toluene
toluene
16
38
54
8
60-79
64
44
59
0
0
0
0
0
-11
-14
70
73
10
3
5
a Used ethyl ester nitroalkene 17 as substrate.
and 2,6-diphenylphenol (83%) could be recovered. As in the
synthesis of (-)-rosmarinecine the intermediate nitronate was
never detected.6
To ensure that the full stereostructure of the nitroso acetal
(+)-29 was correctly assigned, an X-ray crystal structure was
obtained (Figure 4).41 The X-ray analysis verifies the relative
(and absolute) configuration of the six contiguous stereogenic
centers C(2), C(2a), C(7b), C(4a), C(5), and C(6) assuring that
the absolute configurations at the critical stereocenters cor-
respond to those in natural (+)-crotanecine. These configura-
tions are uniquely established by an exo-mode [4 + 2]
cycloaddition followed by an endo-mode42 [3 + 2] cycloaddi-
tion.
Hydrogenolysis of the Nitroso Acetal (+)-29. As was found
in the synthesis of (-)-rosmarinecine,6 rehybridization of C(3)
in (+)-29 was expected to be necessary to unmask the nitroso
acetal to an R-hydroxy lactam. Indeed, that expectation was
borne out as nitroso acetal (+)-29 afforded no identifiable
products under hydrogenolysis with Raney nickel. Therefore,
Figure 4. X-ray crystal structure of nitroso acetal (+)-29 (35%
ellipsoids).
the lactone (+)-29 was reduced with L-Selectride (1.2 equiv)
to afford the lactol (+)-30 in 91% yield (Scheme 4).
The second step in the unmasking procedure was hydro-
genolysis of the nitroso acetal (+)-30 with Raney nickel.
Subjecting the lactol to hydrogenolysis in MeOH afforded the
desired product (-)-31; however, it was accompanied by
numerous side products which were difficult to remove. The
side products were also very unstable and colored rapidly. The
identities of two of the roughly eight to nine different side
products were assigned as the oxazines 37 and 38 from their
spectroscopic data (Figure 5). The hydrogenations in MeOH
showed some pressure dependence with the highest yield being
obtained at 200 psi (Table 4). A change of the solvent to i-PrOH
or n-PrOH afforded slower reaction and lower yield. The results
from reductions in EtOH were comparable to those in MeOH,
except that the side products were not as unstable allowing for
(41) Thorarensen, A.; Swiss, K. A.; Denmark, S. E. Acta Crystallogr. C
1996, 52, 2558.
(42) Defined with respect to folding of the tether.

130 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Denmark and Thorarensen
Scheme 6
Figure 5. Structures of two side products obtained in the hydrogena-
tion.
Table 4. Optimizationa of Hydrogen Pressure in the
Hydrogenolysis of 30
entry
solvent
pressure, psi
yield, %
1
2
3
4
5
6
7
8
MeOH
MeOH
MeOH
EtOH
EtOH
EtOH
160
200
330
150
200
250
280
200
36
54
38
64
44-61
62
b
29
The hydroxyl group at C(1) could be functionalized by
treatment of 32 with methanesulfonyl chloride and triethylamine
to afford the methanesulfonate (+)-33 in excellent yield (95%)
(Scheme 4). Interestingly the R-anomer (+)-32a was mesylated
rapidly while the â-anomer (-)-32b reacted very slowly. This
generally lead to enrichment of the R-anomer during the
functionalization of 32. The difference in the reactivity of the
two anomers is easily understood on the basis of steric repulsion
where the C(2) hydroxyl in the â-anomer is blocked with both
the silyl and methoxy groups inside the concave face of the
tricyclic core. The mesylate (+)-33 was extremely stable and
could be distilled at 250 °C to obtain an analytical sample.
Unmasking of the silyl group to the desired hydroxyl group
was accomplished by the bromination-desilylation-hydroxy-
lation procedure of Fleming.43 Mixing the silane (+)-33 with
potassium bromide and sodium acetate in acetic acid followed
by the addition of peracetic acid afforded (+)-34 in 81% yield
as a white, crystalline compound (Scheme 4). Since the product
was extremely water-soluble, an aqueous workup was avoided.
Therefore, most of the excess peracetic acid was quenched by
stirring with a catalytic amount of 5% Pd/C for 30 min. The
crude product was then purified by silica gel chromatography.
Caution! Under circumstances when the excess peracetic acid
had not been properly quenched a Very exothermic reaction
took place when the residue was applied to silica gel. To obtain
the optimum yield, it was important that the reaction remained
below room temperature during the addition and the destruction
of excess peracetic acid so the mixture was cooled in an ice
bath during these events. This cooling of the reaction mixture
is to minimize hydrolysis of the methyl acetal to a lactol, which
is not compatible with these reaction conditions.
Installation of the Double Bond at C(6)-C(7). The simple
operations of elimination and reduction of lactam (+)-34 were
necessary to complete the synthesis. Since lactam 39 was
readily available from the synthesis of (-)-rosmarinecine, it
served ideally as a model system to probe the installation of
the C(6)-C(7) double bond.6 Treatment of 39 with methane-
sulfonyl chloride and triethylamine afforded the mesylate 40
in 97% yield (Scheme 6). Hydrolysis of the methyl acetal in
90% TFA at room temperature afforded the lactol which was
subsequently treated with several bases (DBU, NaOAc, and K2-
CO3). In all instances the reaction mixture turned black after
the addition of the base. Monitoring the reactions by 1H NMR
spectroscopy showed the disappearance of the lactol, but no
other products could be detected. To probe if this failure was
due to the leaving group, the 4-nitrobenzoate 42 (from the
synthesis of (-)-rosmarinecine) was treated with DBU and the
n-PrOH
i-PrOH
a All reactions were run between 24 and 70 h, and most reactions
were run for 48 h. b Over 50% of the starting material remained after
48 h.
Scheme 5
easier purification. The reaction performed in EtOH showed
very little pressure dependence over the range surveyed, and
all reaction were therefore performed at 200 psi. Under optimal
conditions, the R-hydroxy lactam (+)-31 was isolated in 58%
yield and in 94.9% enantiomeric excess as determined by chiral
HPLC analysis (Scheme 4). A single recrystallization of this
material afforded an 80% recovery of (+)-31 with >99%
enantiomeric excess. In addition, 87% of the chiral auxiliary
(-)-(1R,2S)-2-phenylcyclohexanol was recovered.
Unmasking of the Hydroxyl at C(2). The lactam (+)-31
contains the 1-azabicyclo[3.3.0]octane skeleton of (+)-crotane-
cine, with the correct relative and absolute configuration at C(5),
C(5a), and C(7b) (C(1), C(2), and C(7a) of (+)-crotanecine).
To obtain (+)-crotanecine it remained to unmask the hydroxyl
group at C(2) (crotanecine numbering) and utilize the hydroxyl
group at C(6) to introduce the unsaturation at C(6)-C(7)
(Scheme 5). Finally it was necessary to reduce the lactam and
the lactol to reveal the pyrrolizidinetriol.
To transform the hydroxyl group at C(1) to a good leaving
group, the hydroxyl at C(7) had to be selectively protected. This
was readily accomplished in MeOH at room temperature with
a catalytic amount of TsOH to afford the protected methyl acetal
32 in excellent yield (95%) as a (6/1) mixture of anomers (1H
NMR analysis) (Scheme 4). The major anomer (+)-32a was
assigned the R-configuration on the assumption of thermody-
namic control. In addition, in the 1H NMR spectrum, the HC-
(7) signal appeared as a singlet at 5.24 ppm, indicating a trans
relationship between HC(7) and HC(7a). In the minor â-anomer
(-)-32b HC(7) appeared as a doublet (J ) 6.2 Hz) at 5.12 ppm,
now indicating a cis relationship between HC(7) and HC(7a).
The anomers were separated and characterized. Interestingly
the two compounds have opposite signs of optical rotation; the
R-anomer is dextrorotatory while the â-anomer is levorotatory.
(43) (a) Fleming, I.; Sanderson, P. E. J. Tetrahedron Lett. 1987, 28, 4229.
(b) Fleming, I.; Henning, R.; Parker, D. C.; Plaut, H. E.; Sanderson, P. E.
J. J. Chem. Soc., Perkin Trans. 1 1995, 317.

Synthesis of (+)-Crotanecine
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 131
Treatment of the methyl acetal in 90% TFA for 20 h afforded
only 17% yield of the desired product 35 and 69% recovery of
the methyl acetal. To optimize the hydrolysis of the methyl
acetal, a series of reactions were performed in deuteriotrifluo-
roacetic acid (TFA-d1) diluted to the appropriate concentration
1
with D2O and then monitored by H NMR spectroscopy. All
of the reactions showed first-order kinetics and the linear
correlation was of sufficient quality for the purpose of optimiza-
tion. The reaction performed at room temperature had a half-
life (t1/2) of 37 h. Employing 95% TFA had a marginal effect
(t1/2 35 h), and increasing the temperature to 40 °C led to a t1/2
of 22.7 h. These long reaction times were rather unsatisfactory.
Fortunately, further increase in temperature to 60 °C led to a
more tolerable reaction time with a t1/2 of 5.1 h. Under optimal
reaction conditions the hydrolysis was performed at 60 °C for
24 h. The TFA was then azeotropically removed with benzene,
and the residue was purified on silica gel to afford the lactol
(+)-35 in 75% yield (Scheme 4).51
Treating (+)-35 with large excess of borane (40 equiv) in
THF at reflux reduced both the lactam and the lactol, to afford
the amino triol 36 (Scheme 4). The reaction was cooled to room
temperature and concentrated in vacuo, and the residue was then
dissolved in MeOH and evaporated to dryness repeatedly. For
analytical purposes the amino triol could be purified by silica
gel chromatography in 75% yield. The pyrrolizidinetriol was
of similar polarity to the starting lactam-lactol (+)-35 by TLC,
suggesting that the nitrogen had been temporarily complexed.
It is well-known that borane forms stables adduct with amines,
and indeed, the 11B NMR spectrum showed the presence of a
characteristic resonance for an ate complex at -7.95 ppm.52
The hydrogens attached to borane could not be confidently
identified in the 1H NMR spectrum nor were there any additional
signals in the 13C NMR spectrum. However, the IR spectrum,
where B-H stretches at 2373 and 2337 cm-1 were observed,
allowed the structure to be tentatively assigned as a trihydri-
doborane-pyrrolizidine complex 36.53 Heating a solution of
this compound in triethylamine afforded (+)-crotanecine. Due
to solubility problems it was necessary to perform the reaction
in a mixture of MeOH/Et3N (2/1); more lipophilic alcohols such
as i-PrOH and t-BuOH were not as effective. In addition, it
was necessary to heat the reaction to 120-130 °C for 5 h to
ensure complete consumption of the borane adduct 36. Under
optimal conditions the reaction was performed in a sealed tube
which was degassed under Ar prior to heating. The product
was purified by silica gel chromatography to afford (+)-
crotanecine in 71% yield over the two steps for the reduction-
elimination sequence (Scheme 4).
Figure 6. Various ring structures related to the desired compound 48.
same phenomenon was observed. These results were rather
puzzling, and alternative methods were sought for the construc-
tion of the olefin. A common method for creating unsaturation
at C(6)-C(7) in necines is a selenation at C(7) followed by
oxidative elimination.44 However, the inability to obtain the
desired C(6) selenide thwarted this solution as well.45
Frustrated by the difficulty of this seemingly trivial operation,
we turned to the literature for guidance. Collected in Figure 6
is a group of unsaturated pyrrolizidines, the formation of which
provided useful insights. For example, installation of the double
bond had been accomplished without any difficulties through a
mesylation-elimination sequence to produce 43, and the ring
system in 43 is stable with that degree of unsaturation.46 The
aldehyde lactam in 44 is a stable entity.47 All the required
functionalities are incorporated into 45, which is also a stable
entity.48 Compounds of the structure 46 are unstable and
rearrange readily to 47, indicating that the target compound 48
may not be stable either.49
From this analysis it appeared that all three functional groups
(olefin, lactam, and aldehyde) may not exist simultaneously in
the same ring system. Therefore, we opted to remove the
carbonyl groups sequentially and then install the double bond
to obtain the generic structure 43.
There are several examples of the reduction of a lactam in
the presence of a mesylate,50 but to carry out the desired
reduction of both carbonyl groups, it was first necessary to
deprotect the methyl acetal (+)-34 to the corresponding lactol.
(44) (a) Niwa, H.; Kuroda, A.; Yamada, K. Chem. Lett. 1983, 125. (b)
Niwa, H.; Miyachi, Y.; Okamoto, O.; Uosaki, Y.; Yamada, K. Tetrahedron
Lett. 1986, 27, 4605. (c) Kano, S.; Yuasa, Y.; Shibuya, S. Heterocycles
1988, 27, 253. (d) Robins, D. J.; Sakdarat, S. J. Chem. Soc., Perkin Trans.
1 1979, 1734. (e) Robins, D. J.; Sakdarat, S. J. Chem. Soc., Perkin Trans.
1 1981, 909. (f) Terao, Y.; Imai, N.; Achiwa, K.; Sekiya, M. Chem. Pharm.
Bull. 1982, 30, 3167. (g) Vedejs, E.; Martinez, G. R. J. Am. Chem. Soc.
1980, 102, 7993. (h) Vedejs, E.; Larsen, S.; West, F. G. J. Org. Chem.
1985, 50, 2170.
(45) (a) Liotta, D.; Markiewicz, W.; Santiesteban, H. Tetrahedron Lett.
1977, 4365. (b) Liotta, D.; Santiesteban, H. Tetrahedron Lett. 1977, 4369.
(c) Tiecco, M.; Testaferri, L.; Tingoli, M.; Chianelli, D.; Montanucci, M.
Synth. Commun. 1983, 13, 617. (d) Reich, H. J.; Wollowitz, S.; Trend, J.
E.; Chow, F.; Wendelborn, D. F. J. Org. Chem. 1978, 43, 1697. (e) Liotta,
D.; Sunay, U.; Santiesteban, H.; Markiewicz, W. J. Org. Chem. 1981, 46,
2605. For review, see: (f) Clive, D. L. J. Tetrahedron 1978, 34, 1049. (g)
Monahan, R.; Brown, D.; Waykole, L.; Liotta, D. In Organoselenium
Chemistry; Liotta, D., Ed.; John Wiley: New York, 1987; Chapter 4, p
207.
Purification and Identification of (+)-Crotanecine. The
purification of (+)-crotanecine presented severe difficulties. It
appeared that the side products generated in the solvolysis
promoted the decomposition of (+)-crotanecine, thereby com-
plicating the purification. After silica gel chromatography an
analytically pure sample of (+)-crotanecine was obtained by
recrystallization from EtOH/Et2O which had the same physical
(46) (a) Murray, A.; Proctor, G. R.; Murray, P. J. Tetrahedron Lett. 1995,
36, 291. (b) Murray, A.; Proctor, G. R.; Murray, P. J. Tetrahedron 1996,
52, 3757.
properties (mp 193-197 °C, [R]21 +38.7 (EtOH, c ) 0.52)
D
(47) (a) Sato, T.; Tsujimoto, K.; Matsubayashi, K.-I.; Ishibashi, H.; Ikeda,
M. Chem. Pharm. Bull. 1992, 40, 2308. (b) Ishibashi, H.; Ozeki, H.; Ikeda,
M. J. Chem. Soc., Chem. Commun. 1986, 654.
(48) Keck, G. E.; Cressman, E. N. K.; Enholm, E. J. J. Org. Chem. 1989,
54, 4345.
(51) The resistance of the methyl acetal to hydrolysis was most likely
due to the inductive effect of the additional hydroxyl group at C(5). The
inductive effect disfavors the formation of the intermediate oxocarbenium
ion, which is a necessary intermediate in the hydrolysis.
(49) (a) Cabezas, N.; Thierry, J.; Potier, P. Heterocycles 1989, 28, 607.
Other preparations of 47: (b) Gramain, J.-C.; Remuson, R.; Valle´e, D. J.
Org. Chem. 1985, 50, 710. (c) Pinnick, H. W.; Chang, Y.-H. J. Org. Chem.
1978, 43, 4662.
(50) Casiraghi, G.; Spanu, P.; Rassu, G.; Pinna, L.; Ulgheri, F. J. Org.
Chem. 1994, 59, 2906.
(52) (a) Brown, H. C.; Heim, P.; Yoon, N. M. J. Am. Chem. Soc. 1970,
92, 1637. (b) Brown, H. C.; Heim, P. J. Org. Chem. 1973, 38, 912. (c)
Brown, H. C.; Choi, Y. M.; Narasimhan, S. J. Org. Chem. 1982, 47, 3153.
(53) Pretsch, E.; Seibl, J.; Clere, T.; Simon, W. Tables of Spectral Data
for Structure Determination of Organic Compounds, 2nd ed.; Springer-
Verlag: New York, 1989; p I260.

132 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Denmark and Thorarensen
Table 5. Chemical Shifts of (+)-Crotanecine and the Hydrochloride Salt (49)
1H NMR
13C NMR
49,b ppm (integral)
(+)-1,a ppm (integral)
(+)-1,b ppm (integral)
49,b ppm
(+)-1,a ppm
(+)-1,b ppm
5.76 (1H)
4.86 (1H)
4.43 (1H)
4.37-4.19 (4H)
3.94 (1H)
3.80 (1H)
3.13 (1H)
5.77 (1H)
4.90 (1H)
4.45 (1H)
4.39-4.23 (4H)
3.92 (1H)
3.84 (1H)
3.12 (1H)
5.70 (1H)
4.19-4.15 (4H)
4.03 (1H)
3.85 (1H)
3.41-3.37 (1H)
3.23 (1H)
139.81
121.21
77.59
73.38
70.91
64.11
59.21
57.32
139.5
121.4
77.3
73.2
71.0
63.7
59.2
57.1
140.27
125.27
76.12
74.77
72.33
64.20
59.88
59.30
2.60 (1H)
a From ref 13. b This work.
(lit.9 mp 192 °C, [R]20 +39.2 (EtOH, c ) 1.3))54) as the
of mixing. In addition, the cyanide, in neither case, exists as
free LiCN in solution. The lower order cuprate has been
demonstrated to have low selectivity in addition to monosub-
stituted acetylenes, as was observed for 24.
D
material obtained by hydrolysis of anacrotine (74%) (mp 192-
195 °C, [R]21 +35.0 (EtOH, c ) 0.644)). In addition the
D
spectroscopic profiles (1H NMR, 13C NMR, IR, and MS) of
the synthetic and natural (+)-crotanecine were nearly identical.
Another authentic sample of natural (+)-crotanecine was
The temperature dependence of the regioisomeric addition
products obtained in the silylcupration of 24 is puzzling. Two
explanations are proposed: (1) the â-bound copper species (i)
(Scheme 8) would undergo elimination at higher temperatures
to regenerate the alkoxide along with the silylacetylene as a
byproduct, or (2) the reaction is readily reversible and exclusive
isolation of the â-addition product at higher temperature
represents the thermodynamic ratio between the R- and the
â-bound vinylcopper species. If elimination of the â-vinylcop-
per species occurred at higher temperature, only 66% of the
yield from the low-temperature reaction (2/1 mixture) would
be expected. Therefore, the first scenario can be ruled out on
the basis of the similar yield obtained in the two reactions (-78
°C (69%), 0 °C (85%)).
1
obtained from Australia, and the H and 13C NMR spectra of
this sample matched the spectra of synthetic (+)-crotanecine
and the natural sample from anacrotine (see Supporting Infor-
mation).
(+)-Crotanecine displays the most dramatic chemical shift
1
variances in the H NMR spectrum of the three alkaloids that
have been prepared using the tandem [4 + 2]/[3 + 2]
methodology. The 1H NMR chemical shifts are very sensitive
to traces of salts or hydrates of crotanecine. To demonstrate
this, the hydrochloride salt 49 was prepared by dissolving (()-
crotanecine in saturated HCl in MeOH. The chemical shifts
1
for 49 represent one extreme in the H NMR spectrum, while
those for pure (+)-crotanecine were on the other (Table 5).
Various samples of (+)-crotanecine prepared during the course
of this synthetic effort displayed signals throughout this range.
Thermodynamic control of the reaction selectivity at 0 °C is
more likely. The mechanistic hypothesis can therefore be put
forward that the higher order silylcuprate adds in a syn fashion
across the acetylenic ether in a rapidly reversible reaction
(Scheme 7). The kinetic preference for addition is therefore
the same as that observed in carbocupration of acetylene ethers,
but in this instance, the reversible nature of the higher order
cyanosilylcuprate allows the isolation of 26 exclusively. No-
tably, Fleming has observed that the higher order cyanosilyl-
cuprate adds reversibly to allenes.57 A similar observation has
been documented for the addition of higher order cyanostan-
nylcuprates to acetylenic ethers, where in the presence of an
internal trap (MeOH), the R-addition product was isolated
exclusively.58
[4 + 2] Cycloaddition. The mechanism of the [4 + 2]
cycloaddition is considered to be a stepwise process in which
the first carbon-carbon bond forming reaction is irreversible.59
Since the excess vinyl ether (-)-26 recovered from the reaction
was exclusively of the trans configuration, the irreversibility of
this process is further substantiated.28b,60 The stereochemical
outcome of the tandem process is established in the [4 + 2]
cycloaddition by the creation of the C(4) stereocenter. That
stereocenter then controls the formation of the other stereo-
centers as a consequence of the subsequent [3 + 2] process
wherein the tether folds only in an endo orientation.
1
The hydrogens that showed the most dramatic shifts in the H
NMR spectrum were HC(7a) (4.19-4.15 ppm (1), 4.86 (49))
and HC(3) (2.60 ppm (1), 3.13 ppm (49)). To obtain a
consistent spectroscopic profile, the compounds were loaded
on to a silica gel column in MeOH/Et3N (1/2), and the column
was eluted with CHCl3/MeOH (4/1) followed by CHCl3/MeOH/
NH4OH (10/5/1) to remove the (+)-crotanecine. The only other
high-field 1H NMR spectrum reported of (+)-crotanecine13 must
contain a high proportion of protonated material, since the
reported data deviate only slightly from the 1H NMR spectrum
of the hydrochloride salt of (+)-crotanecine.55 In addition, the
differences between the reported 13C NMR spectrum of (+)-
crotanecine13 and the hydrochloride salt of (+)-crotanecine are
minuscule, but both are very different from the 13C NMR
spectrum of authentic (+)-crotanecine.
Discussion
Silylcupration. Although silylcupration has been an impor-
tant synthetic transformation in organic synthesis, there are no
reports of the addition of silyl cuprates to acetylenic ethers.37
The composition of the various silylcopper reagents has been
explored in a series of elegant spectroscopic studies.56 Using
copper(I) cyanide as the copper source, two distinctly different
non-interconverting cuprates, (PhMe2Si)CuCNLi and (PhMe2-
Si)2CuCNLi2, can be formed depending upon the stoichiometry
The critical stereochemical issues in the [4 + 2] cycloaddition
promoted with MAPh and the new vinyl ether (-)-26 are (1)
(57) (a) Cuadrado, P.; Gonza´lez, A. M.; Pulido, F. J.; Fleming, I.
Tetrahedron Lett. 1988, 29, 1825. (b) Barbero, A.; Cuadrado, P.; Gonza´lez,
A. M.; Pulido, F. J.; Fleming, I. J. Chem. Soc., Perkin Trans 1 1991, 2811.
(58) Cabezas, J. A.; Oehlschlager, A. C. Synthesis 1994, 432.
(59) Denmark, S. E.; Moon, Y.-C.; Cramer, C. J.; Dappen, M. S.;
Senanayake, C. B. W. Tetrahedron 1990, 46, 7373.
(54) Reference 9 has the only reported data of analytically pure (+)-
crotanecine; in other publications, the [R]D of synthetic or natural crotanecine
ranges from +25° 13 to +44° 15 and the mp ranges from 18816 to 202 °C.7
(55) Similar observation have been made by Dr. Russell J. Molyneux
(U.S. Department of Agriculture, Albany, CA) (private communications).
(56) (a) Sharma, S.; Oehlschlager, A. C. Tetrahedron 1989, 45, 557.
(b) Sharma, S.; Oehlschlager, A. C. J. Org. Chem. 1989, 54, 5383.
(60) Denmark, S. E.; Schnute, M. E.; Marcin, L. R.; Thorarensen, A. J.
Org. Chem. 1995, 60, 3205.

Synthesis of (+)-Crotanecine
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 133
Figure 7. Two distinct s-cis and s-trans conformations of vinyl ether (-)-26 located by MOPAC calculations with PM3.
Scheme 7
Figure 8. Proposed approach of the dienophile (-)-26 in the [4 + 2]
cycloaddition to 14 promoted by MAPh.
in shielding the re face of the dienophile. In formulating a
transition structure for this cycloaddition it should be remem-
bered that any differences due to small steric or electronic
preferences between the two ground state conformations can
be overcome when challenged with the demands of the transition
structure leading to product. Only the s-trans conformation of
the vinyl ether exposes the reactive face of the vinyl ether in
the cycloaddition. Therefore, the preferred mode of reaction
of (-)-26 in the [4 + 2] cycloaddition promoted by MAPh is
via the exo mode where the dienophile prefers an s-trans
conformation, thus exposing the si face for approach to the si
face of the nitroalkene as depicted in Figure 8.
the reactive conformation of the vinyl ether, (2) the facial
selectivity of the vinyl ether, and (3) the orientation (exo/endo)
of the dienophile with respect to heterodiene.
MOPAC calculations of the vinyl ether (-)-26 using the PM3
Hamiltonian located two distinct conformations of the vinyl
ether (Figure 7). The lowest in energy was an s-cis conforma-
tion; 1.63 kcal higher was an s-trans conformation. In the s-cis
conformation, the si face of the vinyl ether is shielded by the
phenyl ring of the auxiliary, while in the s-trans conformation,
the re face of the vinyl ether is shielded.61
The X-ray crystal structure of the product nitroso acetal (+)-
29 provided clear answers to some of the issues mentioned
above. The cis relationship of HC(4a) and HC(5) is uniquely
established by an exo-mode orientation of the dienophile with
respect to the heterodiene in the [4 + 2] cycloaddition. In
addition, since the absolute configuration of the auxiliary
(1R,2S)-2-phenylcyclohexanol is known, the configurations of
HC(4a) and HC(5) can both be assigned as S. To obtain the
4aS configuration, the vinyl ether (-)-26 must approach the si
face of the nitroalkene 14.62 Furthermore, to obtain the 5S
configuration, the nitroalkene 14 has to approach the si face of
the vinyl ether. To obtain the observed selectivity, which is
the best recorded so far from an exo-mode cycloaddition of a
2-substituted vinyl ether, the auxiliary has to be very efficient
Mechanistic Implications of the Hydrogenolysis. The
hydrogenolysis of the nitroso acetal is the pivotal operation for
the construction of the 1-azabicyclo[3.3.0]octane ring system
in all of the necine syntheses which employ the tandem [4 +
2]/[3 + 2] cycloaddition. These reactions have required
extensive optimization compared to the high yielding hydro-
genolyses in simpler model systems. In an earlier report, a
detailed discussion was presented for the individual steps in
this process,5
and indeed, the isolation of the intermediates 37
and 38 supports the notion of preferential cleavage of the five-
membered ring.63 Scheme 8 depicts additional problems which
could arise if the initial N-O cleavage occurs in the six-
membered ring. In this scenario, the rate of cleavage of the
second N-O bond in i leading to ii must be faster than both
the elimination and lactamization which form the oxazines 37
and 38. It is apparent that, in this reaction, the rate of
elimination to the oxazines has to be comparable to the rate of
second N-O bond cleavage. Furthermore, breakdown of the
(63) An intermediate resulting from the cleavage of the five-membered
ring has been observed for nitroso acetal derived from the tandem inter[4
+ 2]/inter[3 + 2] cycloaddition sequence employing nickel-aluminum
alloy/sodium hydroxide. Schnute, M. E. Ph.D. Thesis, University of Illinois,
Urbana, 1995, p 139.
(61) The si and re faces of the vinyl ether are defined with respect to
the oxygen.
(62) The si and re faces of the nitroalkene are defined with respect to
the nitrogen.

134 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Denmark and Thorarensen
Scheme 8
differentiates its reactivity from 52 is the contraction of the five-
membered ring lactone in (+)-29 compared to the carbocycle
found in 52. The bonds O-C(4a), O-C(3), and C(3)-C(2a)
are 0.08, 0.18, and 0.03 Å shorter than the corresponding C-C
bonds in 52. Taken together with the sp2 hybridization of C(3),
these changes increase the strain in this ring and may slow the
rate of the second N-O bond cleavage and subsequent imine
formation, thereby allowing possible side reactions to intervene
with (+)-29 compared to 52.
For the continued success of the tandem [4 + 2]/[3 + 2]
cycloaddition method, a thorough understanding of this appar-
ently simple reaction is necessary. In addition it would be
desirable to find an easily handled catalyst, which could be
accurately metered and would be homogeneous under the
reaction conditions to avoid many of the problems that have
been encountered.
Conclusion
The (E)-2-silylvinyl ether (-)-26 has afforded the best
selectivity of any substituted dienophile explored so far in the
exo-mode [4 + 2] cycloaddition promoted by MAPh. Using
this reaction the synthesis of (+)-crotanecine was completed
in 10 steps and 10.2% overall yield. This constitutes the second
example of the fused-mode [4 + 2]/[3 + 2] cycloaddition in
necine base synthesis. It demonstrates that, with this method,
the basic ring system can be functionalized by the employment
of substituted vinyl ethers as dienophiles. In addition, unsat-
uration can be introduced at the C(6)-C(7) position, a common
functionality many necines share. The extension of this
methodology to the synthesis of more structurally challenging
pyrrolizidine alkaloids is under active investigation.
Experimental Section
Scheme 9
General Experimental Procedures. See Supporting Information.
rel-(1R,2S)-trans-2-Phenyl-1-[[2-(dimethylphenylsilyl)ethynyl]-
oxy]cyclohexane (25). A solution of the acetylenic ether 24 (1.040 g,
5.19 mmol) in THF (30 mL) was cooled to -74 °C followed by a
dropwise addition of n-BuLi (1.51 M, 3.61 mL, 5.45 mmol, 1.05 equiv).
The resulting solution was stirred for 20 min, chlorodimethylphenyl-
silane (0.91 mL, 5.45 mmol, 1.05 equiv) was added, and the reaction
mixture was stirred for 1 h at -74 °C, then warmed to 0 °C and stirred
for 1 h. The reaction mixture was diluted with pentane (300 mL),
washed with NaHCO3 (1 × 100 mL) and brine (1 × 100 mL), dried
with Na2SO4 and filtered. The organic extract was concentrated in
vacuo, and the crude product was purified by chromatography on basic
alumina II (pentane/MTBE, (1/0, 18/1, 9/1)) to afford 1.484 g (85%)
of silane 25. An analytical sample was obtained by high-vacuum, bulb-
to-bulb distillation. 25: bp 120-125 °C (3.2 × 10-5 mmHg); 1H NMR
(400 MHz, CDCl3) 7.60-7.58 (m, 2H), 7.36-7.23 (m, 8H), 4.19 (dt,
Jd ) 4.5, Jt ) 10.8, 1H), 2.77 (ddd, J ) 3.8, 10.6, 12.3, 1H), 2.46-
2.42 (m, 1H), 1.97-1.92 (m, 2H), 1.79-1.29 (m, 5H), 0.32 (s, 6H);
13C NMR (100 MHz, CDCl3) 142.15, 138.64, 133.58, 128.91, 128.40,
127.59, 127.53, 126.71, 109.11, 90.16, 49.01, 33.75, 30.88, 25.41, 24.65,
24.62, -0.05; IR (neat) 2170 (s); MS (70 eV) 334 (M+, 0.72); TLC Rf
) 0.46 (hexane/EtOAc, 19/1). Anal. Calcd for C22H26OSi (334.53):
C, 78.98%; H, 7.83%. Found: C, 78.98%; H, 7.85%.
rel-(1R,2S)-trans-2-Phenyl-1-[[(Z)-2-(dimethylphenylsilyl)ethenyl]-
oxy]cyclohexane (27). To a solution of propenyl ether 25 (0.530 g,
1.58 mmol) in THF (12 mL) at room temperature (rt) was added a
solution of diisobutylaluminum hydride (2.18 M in THF, 2.18 mL, 4.75
mmol, 3.0 equiv). The resulting mixture was stirred for 2 h then was
cooled to 0 °C and quenched with water. The mixture was poured
into CH2Cl2 (100 mL) and was washed with water (2 × 50 mL). The
aqueous layer was back-extracted with MTBE (50 mL), and the
combined organic extracts were dried (Na2SO4), filtered, and concen-
trated in vacuo. The residue was purified by chromatography on basic
alumina activity II (pentane/MTBE, (1/0, 4/1, 1/1)) to afford 368 mg
(69%) of 27. An analytical sample was prepared by high-vacuum, bulb-
hemiacetal in ii and imine formation to iV has to be faster than
lactamization and overreduction affording iii for the overall
transformation to be successful.
The hydrogenolysis of the nitroso acetal with the general
structure 50 affords a high yield (70-90%) of the final product
under very simple reaction conditions (Raney nickel/1 atm H2/
MeOH) (Scheme 9).4 In this scenario, the rate of both N-O
bond cleavages and imine formation must be faster than possible
side reactions. How do the structural differences between the
simplified, all-carbon nitroso acetals and the lactone-containing
nitroso acetal 53 influence their behavior toward reduction? The
X-ray crystal structure of (+)-29 allows the comparison of these
two different systems and further comparison to a simpler
analogue of a similar ring structure 52.64 In both of these
structures the tetrahydro-1,2-oxazine ring exists in a twist-boat
conformation. A structural attribute of (+)-29 which critically
(64) Denmark, S. E.; Moon, Y.-C.; Senanayake, C. B. W. J. Am. Chem.
Soc. 1990, 112, 311.

Synthesis of (+)-Crotanecine
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 135
to-bulb distillation. 27: bp 100-110 °C (4 × 10-5 mmHg); 1H NMR
(400 MHz, C6D6) 7.61-7.59 (m, 2H), 7.27-7.07 (m, 8H), 6.27 (d, J
) 8.3, 1H), 4.17 (d, J ) 8.3, 1H), 3.35 (dt, Jd ) 4.4, Jt ) 10.5, 1H),
2.39 (ddd, J ) 3.4, 10.1, 12.2, 1H), 1.85-1.81 (m, 1H), 1.66-1.62
(m, 1H), 1.51-1.48 (m, 1H), 1.41-1.39 (m, 1H), 1.24-1.16 (m, 2H),
1.02-0.93 (m, 2H), 0.43 (s, 3H), 0.40 (s, 3H); 13C NMR (100 MHz,
C6D6) 158.95, 143.80, 140.66, 134.19, 128.73, 128.50, 128.09, 127.83,
127.56, 126.61, 96.50, 84.63, 50.72, 33.70, 33.32, 25.81, 24.90, -0.83,
-1.07; IR (neat) 2857 (m), 1605 (s), 1494 (w); MS (70 eV) 336 (M+,
3). Anal. Calcd for C22H28OSi (336.54): C, 78.51%; H, 8.38%.
Found: C, 78.67%; H, 8.44%.
CDCl3) 7.41-7.15 (m, 10H), 5.23 (d, J ) 3.6, 1H), 5.07 (septet, J )
6.3, 1H), 4.72 (d, J ) 7.6, 1H), 4.42 (dd, J ) 1.5, 6.6, 1H), 4.25 (dd,
J ) 6.6, 8.3, 1H), 4.02-3.96 (m, 1H), 3.72 (dd, J ) 3.5, 8.5, 1H),
2.63-2.57 (m, 1H), 2.48-2.46 (m, 1H), 1.86-1.81 (m, 2H), 1.72-
1.69 (m, 1H), 1.54-1.34 (m, 5H), 1.29 (d, J ) 5.8, 6H), -0.10 (s,
3H), -0.17 (s, 3H); 13C NMR (100 MHz, CDCl3) 173.86, 167.14,
144.89, 135.95, 133.93, 129.46, 128.66, 127.93, 127.76, 126.36, 101.46,
84.86, 79.78, 77.16, 75.17, 70.47, 51.25, 49.27, 36.16, 34.41, 28.98,
25.87, 25.18, 21.58, 21.55, -3.63, -5.53; IR (KBr) 1782 (s), 1745
(s); MS (FAB) 566 (M+ + H, 5); [R]23 + 53.9° (CHCl3, c ) 0.83);
D
TLC Rf ) 0.20 (hexane/EtOAc, 4/1). Anal. Calc for C31H39NO7Si
(565.73): C, 65.81%; H, 6.94; N, 2.47%. Found: C, 66.05%; H,
7.05%; N, 2.47%.
(1R,2S)-trans-2-Phenyl-1-[[(E)-2-(dimethylphenylsilyl)ethenyl]-
oxy]cyclohexane ((-)-26). In a 1-L three-necked, round-bottomed
flask was placed CuCN (5.80 g, 64.8 mmol, 2.0 equiv) and THF (350
mL). The white suspension was cooled to -10 °C followed by the
addition of a solution of lithiodimethylphenylsilane (0.90 M, 144 mL,
129.6 mmol, 4.0 equiv). The resulting black solution was stirred for
30 min. A solution of (-)-24 (6.48 g, 32.4 mmol) in THF (60 mL)
was added to the cuprate over 15 min, and the reaction mixture was
then stirred at 0 °C. After 1.5 h, the mixture was poured into water
(0.7 L) and the biphase was extracted with MTBE (3 × 0.3 L). The
combined organic extracts were filtered through Celite and rewashed
with brine (2 × 0.25 L). The organic layer was dried (Na2SO4), filtered,
and concentrated in vacuo. The crude product was purified by
chromatography on basic alumina II (pentane/MTBE, (1/0, 19/1, 9/1)),
and the mixed fractions were repurified by basic alumina II (pentane/
MTBE, (1/0, 19/1, 9/1)) to afford 9.24 g (85%) of vinyl ether (-)-26.
An analytical sample was obtained by high-vacuum, bulb-to-bulb
(2R,2aS,4aS,5S,6S,7bR)-6-[[(1R,2S)-trans-2-Phenylcyclohexyl]-
oxy]-3-hydroxyoctahydro-5-(dimethylphenylsilyl)-1,4,7-trioxa-7a-
azabicyclopent[cd]indenecarboxylic Acid 1-Methylethyl Ester ((+)-
30). To a solution of (+)-29 (1.555 g, 2.74 mmol) in CH2Cl2 (60 mL)
at -74 °C was added dropwise a solution of lithium tris(sec-butyl)-
borohydride (1.09 M in THF, 3.01 mL, 3.29 mmol, 1.2 equiv). The
resulting solution was stirred for 70 min and was then quenched with
a solution of aqueous phosphate buffer (pH 6.9)/glycerol (1/1, 15 mL).
The mixture was immediately poured into CH2Cl2 (300 mL), then
washed with water (1 × 100 mL) and brine (2 × 100 mL), and back-
extracted with CH2Cl2 (1 × 100 mL). The combined organic extracts
were dried (Na2SO4), filtered, and concentrated in vacuo. The crude
product was purified by silica gel column chromatography eluting with
hexane/EtOAc (4/1, 2/1, 1/1) to afford 1.42 g (91%) of (+)-30 as a
1
white foam, the composition of (+)-30 was determined by H NMR
1
distillation. (-)-26: bp 130 °C (4 × 10-5 mmHg); H NMR (400
of the anomeric proton HC(3) to be (3a)/(3b) 16.3/1. An analytical
sample was prepared by recrystallization from hexane/EtOAc. (+)-
30: mp (sealed tube) 133-136 °C dec (hexane/EtOAc); 1H NMR (400
MHz, CDCl3) 7.39-7.15 (m, 10H), 5.45 (d, J ) 2.0, 0.9H), 5.30 (dd,
J ) 6.8, 12.9, 0.1H), 5.18 (d, J ) 3.4, 0.1H), 5.10 septet, J ) 6.3,
1H), 4.93 (d, J ) 7.1, 0.9H), 4.72 (d, J ) 7.1, 1H), 4.32 (d, J ) 12.9,
0.1H), 4.25-4.18 (m, 2H), 4.04-3.96 (m, 1H), 3.39 (ddd, J ) 3.4,
6.8, 8.5, 0.1H), 3.13 (t, J ) 7.6, 0.9H), 2.61-2.49 (m, 2H), 2.19 (d, J
) 2.7, 0.9H), 1.83-1.78 (m, 2H), 1.69-1.66 (m, 1H), 1.52-1.30 (m,
4H), 1.28 (d, J ) 6.1, 6H), 1.15 (dd, J ) 1.6, 7.2, 1H), -0.08 (s,
0.3H), -0.12 (s, 2.7H), -0.19 (s, 0.3H), -0.22 (s, 2.7H); 13C NMR
(100 MHz, CDCl3) 168.34, 145.23, 136.82, 134.05, 129.11, 128.53,
127.72, 127.53, 126.16, 101.32, 101.11, 84.04, 79.13, 78.41, 75.72,
69.64, 58.61, 51.31, 36.21, 34.48, 27.98, 25.94, 25.23, 21.63, -3.54,
-4.80; IR (KBr) 3543 (m), 1732 (s); MS (FAB) 606 (M+ + K, 3);
[R]23D +56.6° (CHCl3, c ) 1.09); TLC Rf ) 0.13 (hexane/EtOAc, 4/1).
Anal. Calcd for C31H41NO7Si (567.75): C, 65.58%; H, 7.27%; N,
2.46%. Found: C, 65.62%; H, 7.44%; N, 2.38%.
MHz, CDCl3) 7.39-7.36 (m, 2H), 7.32-7.26 (m, 5H), 7.21-7.19 (m,
3H), 5.90 (d, J ) 14.6, 1 H), 4.46 (d, J ) 14.6, 1H), 3.91-3.85 (m,
1H), 2.64 (ddd, J ) 3.7, 10.0, 12.4, 1H), 2.20-2.16 (m, 1H), 1.93-
1.85 (m, 2H), 1.79 (m, 1H), 1.59-1.45 (m, 2H), 1.43-1.31 (m, 2H),
0.14 (s, 3H), 0.14 (s, 3H); 13C NMR (100 MHz, CDCl3) 156.98, 143.45,
139.69, 133.79, 128.61, 128.28, 127.87, 127.50, 126.35, 93.63, 82.53,
50.78, 33.70, 32.57, 25.81, 24.90, -1.85, -1.90; IR (neat) 2935 (s),
1613 (s), 1494 (w); MS (70 eV) 336 (M+, 3); [R]23D -11.6 (CHCl3, c
) 1.06); TLC Rf ) 0.47 (hexane/EtOAc, 19/1). Anal. Calcd for
C22H28OSi (336.54): C, 78.51%; H, 8.38%. Found: C, 78.47%; H,
8.38%.
(2R,2aS,4aS,5S,6S,7bR)-6-[[(1R,2S)-trans-2-Phenylcyclohexyl]-
oxy]-3-oxooctahydro-5-(dimethylphenylsilyl)-1,4,7-trioxa-7a-azabi-
cyclopent[cd]indenecarboxylic Acid 1-Methylethyl Ester ((+)-29).
To a solution of 2,6-diphenylphenol (9.26 g, 37.6 mmol, 10.0 equiv)
in toluene (80 mL) was added trimethylaluminum (2.0 M in toluene,
9.40 mL, 18.8 mmol, 5.0 equiv). Gas evolution was observed, and
the resulting light yellow solution was stirred at rt for 30 min.
(1R,5aS,5R,7aR,7bR)-1,7-Dihydroxy-6-oxa-5-(dimethylphenyl-
silyl)octahydro-2H-cyclopenta[gh]pyrrolizin-2-one ((+)-31), (2S,3S)-
2,4-Dihydroxy-3-[5-(dimethylphenylsilyl)-2H-1,2-oxazine]butano-
ic Acid 1-Methylethyl Ester (37), (2S,3S)-2-Hydroxy-3-[5-(di-
methylphenylsilyl)-2H-1,2-oxazinyl]butyrolactone (38). To a solution
of (+)-30 (910 mg, 1.60 mmol) in EtOH (150 mL) was added a catalytic
amount of EtOH-washed (300 mL) W-2 Raney nickel. The suspension
was stirred at rt in a 300-mL flask inside a steel autoclave for 48 h
under a 200 psi atmosphere of H2. The catalyst was filtered off through
Celite, washed with methanol (200 mL), and then concentrated in vacuo.
The residue was separated by silica gel column chromatography eluting
with hexane/EtOAc (4/1, 2/1, 1/1, 1/2, 0/1, then CHCl3/MeOH 9/1,
6/1) into four fractions. The first fraction contained 245 mg (87%) of
recovered (1R,2S)-trans-2-phenylcyclohexanol. The second fraction
contained 9 mg of the lactone oxazine 38, and the third fractions
contained 47 mg of the isopropyloxazine 37. The oxazines were very
unstable, and multiple purifications were required to isolate them in a
somewhat pure state. The fourth fraction contained 296 mg (58%) of
(+)-31 as a white amorphous solid. The enantiomeric purity of (+)-
31 was determined by chiral HPLC to be 94.9% ee. An analytical
sample of (+)-31 was obtained by recrystallization from EtOAc/hexane.
(+)-31: mp (sealed tube) 125-135 °C dec (hexane/EtOAc); 1H NMR
(400 MHz, CDCl3) 7.44-7.41 (m, 2H), 7.27-7.24 (m, 3H), 5.65 (d, J
) 2.4, 1H), 4.73 (dd, J ) 3.1, 3.8, 1H), 4.60 (d, J ) 8.0, 1H), 4.03
(dd, J ) 3.0, 5.4, 1H), 3.75 (dd, J ) 8.6, 11.6, 1H), 3.44 (s, 1H, OH),
3.09-3.02 (m, 2H), 2.80 (s, 1H), 1.75 (ddd, J ) 4.1, 8.5, 10.8, 1H),
To a solution of 14 (0.862 g, 3.76 mmol) and vinyl ether (-)-26
(3.80 g, 11.29 mmol, 3.0 equiv) in toluene (33 mL) at -30 °C was
added the MAPh solution prepared above slowly over 12 min keeping
the internal temperature below -16 °C. During the addition the solution
changed color from deep, dark brown to light brown. The brown
solution was stirred for additional 3.7 h and then was quenched with
MeOH (1.5 mL), poured into CH2Cl2 (1 L), and washed with water (2
× 0.5 L). The aqueous phases were back-extracted with MTBE (2 ×
0.5 L), and the ether extracts were washed with brine (2 × 0.3 L). The
combined organic extracts were dried (Na2SO4), filtered through Celite,
and concentrated in vacuo. The crude product was purified by silica
gel (330 g) column chromatography eluting with hexane/EtOAc (1/0,
9/1, 6/1, 4/1, 1/1, 0/1). The first fractions contained 2,6-diphenylphenol
(and some vinyl ether), which was recrystallized from hexane (300
mL) to afford 7.70 g (83%) of recovered 2,6-diphenylphenol. The
mother liquor was concentrated in vacuo, and the residue was purified
on basic alumina activity II (pentane/MTBE, (1/0, 19/1, 9/1)) to afford
2.49 g (65%) of recovered vinyl ether (-)-26. Later fractions from
the silica gel chromatography were repurified by silica gel chroma-
tography (hexane/EtOAc, (12/1, 9/1, 6/1, 1/1)) to afford a combined
yield of 1.556 g (73%) of (+)-29 as a white foam, the composition of
1
which was determined by H NMR to be a >50/1 (exo/endo, at 500
MHz) mixture of diastereomers 29. An analytical sample and a X-ray
suitable crystals were obtained by recrystallization from MeOH. (+)-
29: mp (sealed tube) 201-203 °C dec (MeOH); 1H NMR (400 MHz,

136 J. Am. Chem. Soc., Vol. 119, No. 1, 1997
Denmark and Thorarensen
0.29 (s, 3H), 0.23 (s, 3H); 13C NMR (100 MHz, CDCl3) 177.27, 137.86,
133.56, 129.24, 127.88, 98.81, 84.48, 71.83, 67.91, 51.80, 45.30, 30.39,
-3.34, -3.41; IR (KBr) 3543 (w), 1732 (s); MS (CI, CH4) 348 (14),
320 (M+ + H, 31); [R]25D -116.5° (CHCl3, c ) 0.60); TLC Rf ) 0.36
(CHCl3/MeOH, 9/1); HPLC (Column Daicel, Chiralpak AD amylose
tris(3,5-dimethylphenylcarbamate) (25 cm × 4.6 mm); method hexane/
EtOH, 90/10, 0.9 mL/min) tR (1S,5aR,5R,7aS,7bR) 18.1 min (97.4%),
tR (1R,5aS,5S,7aR,7bS) 27.2 min (2.5%). Anal. Calcd for C16H21NO4-
Si (319.42): C, 60.16%; H, 6.62%; N, 4.38%. Found: C, 59.99%; H,
6.77%; N, 4.30%. 37: 1H NMR (400 MHz, CDCl3) 9.04 (s, 1H), 7.55-
7.28 (m, 5H), 6.74 (t, J ) 1.8, 1H), 6.02 (t, J ) 1.9, 1H), 4.95 (septet,
J ) 6.2, 1H), 4.59 (d, J ) 2.7, 1H), 3.98 (d, J ) 6.3, 2H), 3.42 (dt, Jd
) 2.8, Jt ) 6.4, 2H), 1.20 (d, J ) 6.4, 3H), 1.09 (d, J ) 6.1 3H), 0.42
(s, 3H), 0.42 (s, 3H); 13C NMR (100 MHz, CDCl3) 173.07, 140.05,
133.84, 128.57, 128.28, 127.49, 124.87, 114.26, 113.31, 71.66, 70.06,
64.17, 43.38, 21.63, -1.51, -1.58; IR (CDCl3) 3608 (w), 3520 (w),
1725 (s); MS (70 eV) 378 (M+ + H, 2); HRMS calcd for (C19H28-
NO5Si) 378.172095, found 378.173677. 38: 1H NMR (400 MHz,
CDCl3) 8.67 (s, 1H), 7.56-7.53 (m, 2H), 7.36-7.32 (m, 3H), 6.83
(dd, J ) 1.5, 2.2, 1H), 6.03 (m, 1H), 4.66 (dd, J ) 8.2, 8.9, 1H), 4.39
(d, J ) 10.9, 1H), 4.34 (dd, J ) 9.0, 11.0, 1H), 3.71 (dddd, J ) 0.7,
8.1, 11.0, 11.0, 1H), 0.46 (s, 6H); MS (70 eV) 318 (M+ + H, 1).
(4/1, 2/1, 1/1, 1/2, 0/1) to afford 505 mg (95%) of (+)-33 as a 11/1
mixture of methyl anomers as determined by 1H NMR. An analytical
sample was prepared by high-vacuum, bulb-to-bulb distillation. (+)-
33: bp 250 °C (4.8 × 10-5 mmHg); 1H NMR (400 MHz, CDCl3) 7.55-
7.53 (m, 2H), 7.38-7.34 (m, 3H), 5.56 (d, J ) 8.5, 1H), 5.16 (s, 1H),
4.61 (t, J ) 3.5, 1H), 4.10 (dd, J ) 3.0, 5.6, 1H), 4.88 (dd, J ) 8.4,
11.6, 1H), 3.32 (s, 3H), 3.28 (s, 3H), 3.24-3.18 (m, 2H), 1.85 (ddd, J
) 4.1, 8.2, 10.6, 1H), 0.42 (s, 3H), 0.37 (s, 3H); 13C NMR (100 MHz,
CDCl3) 170.85, 137.50, 133.56, 129.32, 127.86, 105.21, 84.30, 78.11,
67.17, 54.64, 50.12, 45.74, 39.98, 30.10, -3.31, -3.50; IR (CHCl3)
1721 (s); MS (CI, CH4) 414 (11), 413 (25), 412 (M+ + H, 100); [R]25
D
+134.1° (CHCl3, c ) 0.57); TLC Rf ) 0.77 (CHCl3/MeOH, 9/1). Anal.
Calcd for C18H25NO6SiS (411.54): C, 52.53%; H, 6.12%; N, 3.40%.
Found: C, 52.38%; H, 6.16%; N, 3.30%.
(1R,5aS,5R,7aR,7bR)-5-Hydroxy-1-(methanesulfonyloxy)-7-meth-
oxy-6-oxaoctahydro-2H-cyclopenta[gh]pyrrolizin-2-one ((+)-34). A
solution of (+)-33 (0.501 g, 1.217 mmol), potassium bromide (0.173
g, 1.46 mmol, 1.2 equiv), and sodium acetate (1.23 g, 15.0 mmol, 12.4
equiv) in acetic acid (2.9 mL) was cooled to 0 °C. Peracetic acid (8.9
mL, 36% v/v) was added dropwise at a rate to maintain the reaction
temperature below rt (ca. 20 min). Gas evolution was observed during
the addition of the peracetic acid. The resulting solution was stirred
at rt for 3 h then was cooled to 0 °C. The reaction mixture was
quenched by adding a catalytic amount of 5% Pd/C, and the resulting
black suspension was stirred at 0 °C for 30 min and then was directly
purified by silica gel chromatography eluting with CHCl3/MeOH (1/0,
19/1, 9/1). (Note: extreme care should be taken handling the peracetic
acid residue during the silica gel chromatography, since a Violent
exothermic reaction can take place if the reaction is not quenched
properly.) The crude product was repurified by silica gel chromatog-
raphy eluting with CHCl3/MeOH (1/0, 19/1, 9/1) to afford 291 mg
(81%) of (+)-34. An analytical sample was obtained by recrystalli-
zation from acetone/Et2O/pentane. (+)-34: mp (sealed tube) 145 °C
(1R,5aS,5R,7R,7aR,7bR) and (1R,5aS,5R,7S,7aR,7bR)-1-Hydroxy-
7-methoxy-6-oxa-5-(dimethylphenylsilyl)octahydro-2H-cyclopenta-
[gh]pyrrolizin-2-one (32). To a solution of (+)-31 (0.445 g, 1.39
mmol) and p-toluenesulfonic acid (0.072 g, 0.422 mmol, 0.30 equiv)
in MeOH (85 mL) was added trimethyl orthoformate (3.3 mL). The
resulting solution was stirred at rt for 13 h and was quenched with
poly(vinylpyridine) (0.089 g). The suspension was filtered, and the
filtrate was concentrated in vacuo. The resulting crude product was
purified by basic alumina (activity II) column chromatography eluting
with CHCl3/MeOH (1/1, 6/1, 4/1, 2/1) to afford 0.439 g (95%) of 32
(mixture of methyl acetal anomer) as a light brown oil. The composi-
1
1
dec (acetone/Et2O/pentane); H NMR (400 MHz, CD3OD) 5.72 (d, J
tion of 32 was determined by H NMR to be an 6.0/1 mixture of 7R/
) 8.0, 1H), 5.24 (s, 1H), 4.66 (dt, Jd ) 4.0, Jt ) 7.3, 1H), 4.44 (dd, J
) 2.8, 4.0, 1H), 4.36 (dd, J ) 2.7, 5.3, 1H), 3.43 (d, J ) 7.6, 2H),
3.36 (s, 3H), 3.27 (s, 3H), 3.28-3.23 (m, 1H); 13C NMR (100 MHz,
CD3OD) 172.84, 107.48, 83.34, 78.25, 73.35, 67.00, 55.53, 52.75, 50.63,
7â anomers. For analytical purposes the anomers were separated by
silica gel chromatography eluting with hexane/EtOAc (2/1, 1/1, 1/2,
1/4, 0/1, then CHCl3/MeOH 4/1) followed by high-vacuum bulb-to-
1
bulb distillation. (+)-32a: (7R) bp 195 °C (5 × 10-5 mmHg); H
39.00; IR (KBr) 1692 (s); MS (CI, CH4) 294 (M+ + H, 24); [R]23
NMR (400 MHz, CDCl3) 7.56-7.54 (m, 2H), 7.37-7.39 (m, 3H), 5.24
(s, 1H), 4.70 (d, J ) 8.3, 1H), 4.58 (t, J ) 3.5, 1H), 4.08 (dd, J ) 2.9,
5.4, 1H), 3.88 (dd, J ) 8.4, 11.6, 1H), 3.51 (s, 1H), 3.28 (s, 3H), 3.21-
3.15 (m, 1H), 3.10 (dd, J ) 5.4, 8.3, 1H), 1.83 (ddd, J ) 4.1, 8.3,
10.6, 1H), 0.41 (s, 3H), 0.36 (s, 3H); 13C NMR (100 MHz, CDCl3)
177.62, 137.86, 133.59, 129.19, 127.79, 105.12, 84.01, 71.65, 67.59,
54.60, 51.01, 45.26, 30.25, -3.32, -3.47; IR (CHCl3) 3528 (w), 1702
(s); MS (CI, CH4) 362 (17), 334 (M+ + H, 33); [R]25D +172.2° (CHCl3,
c ) 0.73); TLC Rf ) 0.56 (CHCl3/MeOH, 9/1), Rf ) 0.33 (EtOAc).
Anal. Calcd for C17H23NO4Si (333.45): C, 61.23%; H, 6.95%; N,
4.20%. Found: C, 61.07%; H, 7.07%; N, 4.18%. (-)-32b: (7â) bp
D
+93.9° (MeOH, c ) 0.46); TLC Rf ) 0.39 (CHCl3/MeOH, 9/1). Anal.
Calcd for C10H15NO7S (293.29): C, 40.95%; H, 5.15%; N, 4.77%.
Found: C, 40.76%; H, 5.21%; N, 4.94%.
(1R,5aS,5R,7S,7aR,7bR)-5-Dihydroxy-1-(methanesulfonyloxy)-6-
oxaoctahydro-2H-cyclopenta[gh]pyrrolizin-2-one ((+)-35). A solu-
tion of (+)-34 (0.304 g, 1.036 mmol) in 90% aqueous trifluoroacetic
acid (TFA) (25 mL) was stirred at 60 °C for 24 h. The solution was
then repeatedly diluted with benzene (3 × 30 mL) and concentrated in
vacuo. The crude product was then purified by silica gel column
chromatography eluting with hexane/EtOAc (2/1, 1/1, then acetone)
and repurified by silica gel chromatography eluting with hexane/EtOAc
(2/1, 1/1, then CHCl3/MeOH 9/1, then acetone) to afford 0.218 g (75%)
1
195 °C (3 × 10-5 mmHg); H NMR (400 MHz, CDCl3) 7.55-7.52
(m, 2H), 7.37-7.34 (m, 3H), 5.12 (d, J ) 6.2, 1H), 4.73 (dd, J ) 6.1,
11.3, 1H), 4.61 (t, J ) 3.7, 1H), 3.99 (dd, J ) 3.8, 4.8, 1H), 3.93 (d,
J ) 11.7, 1H), 3.71 (t, J ) 10.8, 1H), 3.39-3.31 (m, 1H), 3.34 (s,
3H), 3.29-3.25 (m, 1H), 1.81 (dt, Jd ) 3.8, Jt ) 10.2, 1H), 0.43 (s,
3H), 0.39 (s, 3H); 13C NMR (100 MHz, CDCl3) 177.22, 137.42, 133.57,
129.32, 127.91, 108.85, 85.27, 75.24, 66.61, 57.33, 49.37, 45.98, 33.40,
-3.43, -3.46; IR (CHCl3) 1713 (s); MS (CI, CH4) 363 (10), 362 (39),
1
of (+)-35. (+)-35: mp 101-106 °C (EtO2/MeOH); H NMR (400
MHz, CD3OD) 5.73 (d, J ) 7.8, 1H), 5.66 (s, 1H), 4.64 (ddd, J ) 4.3,
6.6, 7.7, 1H), 4.57 (dd, J ) 3.0, 4.2, 1H), 4.37 (dd, J ) 2.9, 5.4, 1H).
3.43-3.40 (m, 2H), 3.27 (s, 3H), 3.22 (ddd, J ) 1.2, 7.8, 5.3, 1H); 13
C
NMR (100 MHz, CD3OD) 172.90, 100.74, 83.03, 78.62, 73.35, 67.20,
53.49, 50.71, 38.99; IR (KBr) 3450 (s, br), 1715 (s); MS (CI, CH4)
280 (M+ + H, 12); HRMS calcd for (C9H14NO7S) 280.049 099, found
280.048 001; [R]23D +67.5° (MeOH, c ) 0.58); TLC Rf ) 0.19 (CHCl3/
MeOH, 9/1).
(1S,2R,6R,7S,7aR)-7-(Hydroxymethyl)-6-(methanesulfonyloxy)-
4-(trihydridoboryl)hexahydro-1H-pyrrolizine-1,2-diol (36). To a
solution of lactol (+)-35 (20 mg, 0.0716 mmol) in THF (20 mL) was
added BH3 (1.0 M in THF, 2.86 mL, 2.86 mmol, 40 equiv). Gas
evolution was observed, and the resulting solution was heated to 72
°C (external 80 °C) for 4 h, whence a white precipitate formed. The
mixture was cooled to rt and concentrated in vacuo. The residual oil
was redissolved in MeOH (2 × 10 mL) and concentrated in vacuo to
afford 23 mg of 36. The crude product was purified by silica gel
chromatography eluting with CHCl3/MeOH (1/0, 19/1, 9/1, 6/1, 4/1)
to give 15.2 mg (75%) of 36. 36: 1H NMR (400 MHz, CD3OD) 5.31
(ddd, J ) 4.4, 6.9, 7.8, 1H), 4.39 (ddd, J ) 3.7, 7.8, 9.7, 1H), 4.12-
334 (M+ + H, 17); [R]23 -21.0° (CHCl3, c ) 0.32); TLC Rf ) 0.59
D
(CHCl3/MeOH, 9/1), Rf ) 0.14 (EtOAc). Anal. Calcd for C17H23-
NO4Si (333.45): C, 61.23%; H, 6.95%; N, 4.20%. Found: C, 61.38%;
H, 7.01%; N, 4.32%.
(1R,5aS,5R,7aR,7bR)-5-(Dimethylphenylsilyl)-1-(methane-
sulfonyloxy)-7-methoxy-6-oxooctahydro-2H-cyclopenta[gh]pyrrolizin-
2-one ((+)-33). To a solution of methyl acetal 32 (433 mg, 1.29 mmol)
in CH2Cl2 (36 mL) was added methanesulfonyl chloride (0.20 mL, 2.59
mmol, 2.0 equiv) and triethylamine (0.36 mL, 2.59 mmol, 2.0 equiv)
at rt. The resulting solution was stirred for 1.5 h and then was diluted
with CH2Cl2 (100 mL) and washed with saturated aqueous CuSO4
solution (50 mL) and brine (50 mL). The aqueous layers were back-
extracted with CH2Cl2 (50 mL), and the combined organic extracts were
dried (Na2SO4), filtered, and concentrated in vacuo. The crude product
was purified by silica gel chromatography eluting with hexane/EtOAc

Synthesis of (+)-Crotanecine
J. Am. Chem. Soc., Vol. 119, No. 1, 1997 137
4.08 (m, 2H), 3.96 (dd, J ) 6.9, 11.0, 1H), 3.75 (dd, J ) 6.8, 12.4,
1H), 3.70 (dd, J ) 3.6, 9.4, 1H), 3.57-3.53 (m, 1.5H, HC(5 and THF),
3.46 (dd, J ) 7.8, 10.8, 1H), 3.22-3.10 (m, 2H), 3.09 (s, 3H), 1.59-
1.28 (m, br, 4H, (BH3 and THF)); 13C NMR (100 MHz, CD3OD) 81.27,
78.27, 73.48, 73.15, 72.15, 67.74, 62.79 (CH2/THF), 58.20, 45.61,
37.81, 30.14 (CH2/THF)); 11B NMR (96 MHz, CD3OD) -7.95; IR
(KBr) 3435 (m, br), 2373 (m), 2337 (m); MS (70 eV) 359 (1), 358 (4),
357 (2), 273 (2), 272 (1), 185 (M+ - MeSO3H, 1); TLC Rf ) 0.20
(CH3Cl/MeOH, 9/1).
gel) at the bottom) eluting with (CHCl3/MeOH/NH4OH, 10/5/1) to give
18.9 mg of a yellow solid. The product was further purified by
chromatography on silica gel (prepared as above, the compound was
loaded on the column in 2/1 Et3N/MeOH) eluting with (CHCl3/MeOH,
4/1, followed by CHCl3/MeOH/NH4OH, 10/5/1) to give 18.5 mg (74%)
of (+)-1 as an off white solid (+)-1: mp 192-195 °C; 1H NMR (500
MHz, CD3OD) 5.70 (d, J ) 1.4, 1H), 4.20-4.16 (m, 4H), 4.05 (t, J )
3.6, 1H), 3.88 (dt, Jd ) 14.6, Jt ) 1.8, 1H), 3.44-3.40 (m, 1H), 3.26
(dd, J ) 6.7, 8.7, 1H), 2.63 (t, J ) 9.6, 1H); 13C NMR (125 MHz,
CD3OD) 140.24, 125.06, 76.20, 74.70, 72.26, 64.19, 59.84, 59.19; IR
(KBr) 3326 (s), 3270 (m), 3265 (m), 3256 (m), 3105 (w); MS (70 eV)
171 (M+, 34); [R]21D + 35.0° (EtOH, c ) 0.64); TLC Rf ) 0.20 (CHCl3/
MeOH/NH4OH 10/5/1).
Synthetic (+)-Crotanecine (1). To a solution of lactol (+)-35 (178
mg, 0.637 mmol) in THF (130 mL) was added BH3 (1.0 M in THF,
25.5 mL, 25.5 mmol, 40 equiv). Gas evolution was observed, and the
resulting solution was heated to 72 °C (external 80 °C) for 4 h, during
which a white precipitate formed. The mixture was cooled to rt and
concentrated in vacuo. The residual oil was redissolved in MeOH (2
× 50 mL) and concentrated in vacuo to afford 200 mg of 36. The
triol was dissolved in MeOH (28 mL) and transferred into a Carius
tube, and Et3N (57 mL) was added. The mixture was degassed by
two freeze/thaw cycles with argon. The clear, colorless reaction
solution was then heated at 120-130 °C for 5 h (whence the solution
turned brown), then was cooled to rt and concentrated in vacuo to ca.
3 mL. The crude product was then purified by chromatography on
silica gel (CHCl3/MeOH, 4/1, then CHCl3/MeOH/NH4OH, 10/5/1) to
afford 77.1 mg (70.7%) of (+)-1 as a light brown solid. An analytical
sample was obtained by recrystallization from EtOH/Et2O. (+)-1: mp
193-197 °C (EtOH/Et2O); 1H NMR (500 MHz, CD3OD) 5.70 (d, J )
1.5, 1H), 4.19-4.15 (m, 4H), 4.03 (t, J ) 3.7, 1H), 3.85 (dt, Jd )
15.0, Jt ) 1.8, 1H), 3.41-3.37 (m, 1H), 3.23 (dd, J ) 6.9, 8.7, 1H),
2.60 (t, J ) 9.5, 1H); 13C NMR (125 MHz, CD3OD) 140.27, 125.27,
76.12, 74.77, 72.33, 64.20, 59.88, 59.30; IR (KBr) 3325 (s), 3277 (s),
3269 (s), 3261 (s), 3248 (s); MS (70 eV) 171 (M+, 30); [R]21D + 38.7°
(EtOH, c ) 0.52); TLC Rf ) 0.20 (CHCl3/MeOH/NH4OH, 10/5/1).
Anal. Calcd for C8H13NO3 (171.19): C, 56.12%; H, 7.65%; N, 8.18%.
Found: C, 55.84%; H, 7.52%; N, 7.95%.
Acknowledgment. We are grateful to the National Institutes
of Health (GM-30938) for generous financial support. A.T.
thanks the University of Illinois and The Icelandic Research
Council (NATO) for Graduate Fellowships. We would like to
thank Prof. Jin K. Cha (University of Alabama, Tuscaloosa)
and Prof. Richard H. Wightman (Heriott-Watt University,
Edinburgh) for additional information regarding their syntheses.
We gratefully acknowledge the gift of (+)-crotanecine from Dr.
Kim Gaul (CSIRO Division of Animal Health, Australia) and
gifts of anacrotine from Dr. Russell J. Molyneux (U.S. Depart-
ment of Agriculture, Albany, California) and Prof. E. Ro¨der
(Pharmazeutisches Institut, Bonn, Germany).
Supporting Information Available: Complete listing of 1H
NMR and 13C NMR data with assignments, infrared absor-
bances, and mass spectrum fragments of all compounds, general
experimental procedures and preparation and spectroscopic data
for compounds 18, 40, and 49 along with 1H NMR, 13C NMR,
IR, and MS spectra of natural and synthetic (+)-crotanecine,
Natural (+)-Crotanecine (1). A solution of anacrotine (4) (0.052
g, 0.147 mmol) and NaOH (40.0 mg, 1.0 mmol, 6.7 equiv) in water
(1.1 mL) was heated at reflux for 2.5 h. The resulting yellow solution
was cooled to rt, diluted with EtOH, and concentrated in vacuo. The
crude product was purified by silica gel chromatography (the column
was prepared with a large Celite plug (one-half the height of the silica
1
and a H NMR comparison of (+)-crotanecine and the hydro-
chloride salt (33 pages). See any current masthead page for
ordering or Internet access instructions.
JA963084D