Application of a stereospecific intramolecular allenylsilane imino ene reaction to enantioselective total synthesis of the 5,11- methanomorphanthridine class of Amaryllidaceae alkaloids

Article abstract of DOI:10.1021/ja970839n

Enantioselective total syntheses of the pentacyclic 5,11- methanomorphanthridine Amaryllidaceae alkaloids (-)-montanine (1), (-)- coccinine (2), and (-)-pancracine (3) were accomplished using an intramolecular concerted pericyclic allenylsilane imino erie

Full text of DOI:10.1021/ja970839n

VOLUME 119, NUMBER 25
J UNE 25, 1997
© Copyright 1997 by the
American Chemical Society
Application of a Stereospecific Intramolecular Allenylsilane
Imino Ene Reaction to Enantioselective Total Synthesis of the
5,11-Methanomorphanthridine Class of Amaryllidaceae
Alkaloids
Jian Jin and Steven M. Weinreb*
Contribution from the Department of Chemistry, The PennsylVania State UniVersity,
UniVersity Park, PennsylVania 16802
ReceiVed March 14, 1997X
Abstract: Enantioselective total syntheses of the pentacyclic 5,11-methanomorphanthridine Amaryllidaceae alkaloids
(-)-montanine (1), (-)-coccinine (2), and (-)-pancracine (3) were accomplished using an intramolecular concerted
pericyclic allenylsilane imino ene cycloaddition as a key step. These complex natural products were constructed
starting from readily available enantiomerically pure epoxy alcohol 15 which was converted to allenylsilane aldehyde
28 Via an efficient nine-step sequence. The imine generated from aldehyde 28 and iminophosphorane 47 underwent
a stereospecific thermal imino ene reaction to afford key intermediate cis aminoalkyne 49. It was possible to transform
this compound Via Lindlar hydrogenation followed by an intramolecular Heck reaction to seven-membered ring
tetracycle 51. This olefinic intermediate could be functionalized through its epoxide to yield R-hydroxymethyl
intermediate 54, and then pentacyclic alcohol 64. Procedures were then developed to convert this material to the
enantiomerically pure alkaloids 1-3. A formal enantioselective total synthesis of (-)-brunsvigine (4) was also
achieved Via triol 72.
Introduction and Background
amarylloides, etc.) collected in South Africa.3 Shortly thereafter,
(-)-brunsvigine (4) was isolated from BrunsVigia cooperi Baker
and BrunsVigia radulosa Herb.4,5 (-)-Pancracine (3) was found
as a minor alkaloid in Rhodophiala bifida, a plant which is
indigenous to the United States, along with (-)-montanine as
the major alkaloid.6
The structural assignments of the 5,11-methanomorphanthri-
dine alkaloids were initially based on chemical degradations
and interconversions.5 In 1968, a spectroscopic study of (-)-
pancracine (3) and some of its derivatives involving proton
NMR and mass spectrometry confirmed that alkaloids 1-5
For decades chemists have been attracted by the Amarylli-
daceae alkaloids due to their diverse and interesting structures.1
The 5,11-methanomorphanthridines constitute one of eight
classes within this large family of alkaloidal natural products.2
Three members of the class, (-)-montanine (1), (-)-coccinine
(2), and (-)-manthine (5), were first isolated in 1955 by
Wildman and co-workers from various Haemanthus species
(Haemanthus montanus, Haemanthus coccineus, Haemanthus
X Abstract published in AdVance ACS Abstracts, June 15, 1997.
(1) For recent reviews, see: (a) Martin, S. F. The Amaryllidaceae
Alkaloids. In The Alkaloids; Brossi, A., Ed.; Academic Press: San Diego,
1987; Vol. 30, p 252. (b) Lewis, J. R. Nat. Prod. Rep. 1993, 10, 291.
(2) For this classification system, see: Dalton, D. R. The Alkaloids;
Marcel Dekker: New York, 1979; p 197.
(3) Wildman, W. C.; Kaufman, C. J. J. Am. Chem. Soc. 1955, 77, 1248.
(4) Dry, L. J.; Poynton, M. E.; Thompson, M. E.; Warren, F. L. J. Chem.
Soc. 1958, 4701.
(5) Inubushi, Y.; Fales, H. M.; Warnhoff, E. W.; Wildman, W. C. J.
Org. Chem. 1960, 25, 2153.
(6) Wildman, W. C.; Brown, C. L. J. Am. Chem. Soc. 1968, 90, 6439.
S0002-7863(97)00839-1 CCC: $14.00 © 1997 American Chemical Society

5774 J. Am. Chem. Soc., Vol. 119, No. 25, 1997
Jin and Weinreb
indeed have the structures previously attributed to them.6 The
structure of (-)-brunsvigine (4) is firmly based on single crystal
X-ray analysis of the bis-p-bromobenzoate derivative and its
absolute configuration was determined by anomalous dispersion
methodology.7 In general, this group of alkaloids all possess a
common bridged pentacyclic skeleton, varying only in the
substitution (i.e., methoxyl or hydroxyl) and stereochemistry
at C-2 and C-3. Interestingly, a recent report has described the
isolation of (+)-montabuphine (6), an alkaloid which is appar-
ently in the enantiomeric series.8,9
total synthesis of the marine alkaloid papuamine.13a,b In
subsequent papers we further demonstrated the scope and
generality of this type of concerted pericyclic process. In
particular, we hoped to apply a stereoselective ene cyclization
of the type previously used to generate cyclohexyl systems with
adjacent cis-amino and -alkynyl moieties (eq 1) to the synthesis
of the montanine group of Amaryllidaceae alkaloids.
Our general strategy for enantioselective synthesis of the 5,11-
methanomorphanthridine alkaloids is outlined in Scheme 1.16
A key compound is the pentacyclic amine 7, closely related to
an intermediate of Hoshino in which the C-2/C-3 oxygens were
not differentially protected, but which could be converted to
racemic alkaloids 1-4.12 The intent was to install the requisite
C-1/C-11a double bond at a late stage of the synthetic exercise.
We anticipated preparing 7 from hydroxymethyl compound 8
Via a transannular cyclization of the type used by the Hoshino
group.12 Intermediate 8 was to be generated by hydroboration
from the least hindered face of exocyclic olefin 9.
Unlike the two previous montanine alkaloid total synthe-
ses11,12 which both used a Pictet-Spengler-type cyclization to
create the C-6/C-6a bond connection, we planned to construct
tetracyclic intermediate 9 by an intramolecular Heck cyclization
of tricyclic bromoalkene 10, derived from alkyne 11, generating
the C-10a/C-11 bond. Although the use of an intramolecular
Heck cyclization to construct a seven-membered ring was
precedented,17 a comparable nitrogen-tethered case had not yet
been reported when our synthesis was initiated.18 We intended
to construct alkyne 11 utilizing our intramolecular pericyclic
imino ene chemistry of enantiomerically pure allenylsilane imine
12 (cf. eq 1) which would be generated by condensation of
scalemic allenylsilane aldehyde 13 and a substituted piper-
onylamine derivative (14).
These Amaryllidaceae alkaloids display some limited biologi-
cal activity. For example, (-)-coccinine (2) shows convulsive
action in high doses [LD50 ) 17.5 mg/kg (in ViVo, dog)].10 Weak
hypotensive and convulsive activities are also reported for (-)-
montanine (1) [LD50 ) 42 mg/kg (in ViVo, dog)]. It might be
relevant that both physiologically active alkaloids have methyl
ether functionality at the C-2 position.
Rather surprisingly, although there has been extensive
synthetic effort in the area of Amaryllidaceae alkaloids,1
construction of the 5,11-methanomorphanthridines has received
little attention from synthetic organic chemists. However, two
groups have recently described successful approaches to total
synthesis of the montanine-type alkaloids. In 1991, Overman
and Shim described total syntheses of both racemic and (-)-
pancracine (3) utilizing a clever tandem aza-Cope rearrange-
ment/Mannich cyclization as the pivotal step.11 In a series of
papers, Hoshino and co-workers reported two elegant strategies
for total syntheses of these alkaloids, resulting in preparation
of 1-4 in racemic form.12
(13) (a) Borzilleri, R. M.; Weinreb, S. M.; Parvez, M. J. Am. Chem.
Soc. 1994, 116, 9789. (b) Borzilleri, R. M.; Weinreb, S. M.; Parvez, M. J.
Am. Chem. Soc. 1995, 117, 10905. (c) Jin, J.; Smith, D. T.; Weinreb, S. M.
J. Org. Chem. 1995, 60, 5366. (d) Weinreb, S. M. J. Heterocycl. Chem.
1996, 33, 1429. (e) Weinreb, S. M.; Smith, D. T.; Jin, J. Synthesis, in press.
For a review of imino ene reactions, see: Borzilleri, R. M.; Weinreb, S.
M. Synthesis 1995, 347.
(14) For intermolecular reactions of allenylsilanes with electrophiles Via
a non-pericyclic process, see: (a) Danheiser, R. L.; Carini, D. J. J. Org.
Chem. 1980, 45, 3925. (b) Danheiser, R. L.; Carini, D. J.; Kwasigroch, C.
A. J. Org. Chem. 1986, 51, 3870. (c) Danheiser, R. L.; Kwasigroch, C. A.;
Tsai, Y.-M. J. Am. Chem. Soc. 1985, 107, 7233. (d) Danheiser, R. L.; Stoner,
E. J.; Koyama, H.; Yamashita, D. S.; Klade, C. A. J. Am. Chem. Soc. 1989,
111, 4407. (e) Becker, D. A.; Danheiser, R. L. J. Am. Chem. Soc. 1989,
111, 389.
(15) For intramolecular metalloene reactions of allenylsilanes, see: (a)
Meyer, C.; Marek, I.; Courtemanche, G.; Normant, J. F. J. Org. Chem.
1995, 60, 863. (b) Meyer, C.; Marek, I.; Normant, J. F. Tetrahedron Lett.
1996, 37, 857. (c) Lorthiois, E.; Marek, I.; Normant, J. F. Tetrahedron Lett.
1997, 38, 89.
(16) (a) A preliminary account of portions of this work has appeared:
Jin, J.; Weinreb, S. M. J. Am. Chem. Soc. 1997, 119, 2050. (b) Jin, J. Ph.D.
Thesis, The Pennsylvania State University, University Park, PA, 1997.
(17) (a) Heck, R. F. Org. React. 1982, 27, 345. (b) Abelman, M. M.;
Oh, T.; Overman, L. E. J. Org. Chem. 1987, 52, 4130.
(18) Cf. (a) Mohanakrishnan, A. K.; Srinivasan, P. C. Tetrahedron Lett.
1996, 37, 2659. (b) Cornec, O.; Joseph, B.; Merour, J. Tetrahedron Lett.
1995, 36, 8587.
Retrosynthetic Plan
In 1994 we described the discovery of a novel intramolecular
imino ene reaction of an allenylsilane and its application to the
(7) Laing, M.; Clark, R. C. Tetrahedron Lett. 1974, 583.
(8) Viladomat, F.; Bastida, J.; Codina, C.; Campbell, W. E.; Mathee, S.
Phytochemistry 1995, 40, 307.
(9) The biosynthesis of these alkaloids has been studied: (a) Fuganti,
C.; Ghiringhelli, D.; Grasselli, P. J. Chem. Soc., Chem. Commun. 1973,
430. (b) Wildman, W. C.; Olesen, B. Ibid. 1976, 551.
(10) Southon, I. W.; Buckingham, J. Dictionary of the Alkaloids;
Chapman & Hall: New York, 1989; pp 229, 735.
(11) (a) Overman, L. E.; Shim, J. J. Org. Chem. 1991, 56, 5005. (b)
Overman, L. E.; Shim, J. J. Org. Chem. 1993, 58, 4662.
(12) (a) Ishizaki, M.; Hoshino, O.; Iitaka Y. J. Org. Chem. 1992, 57,
7285. (b) Ishizaki, M.; Kurihara, K.; Tanazawa, E.; Hoshino, O. J. Chem.
Soc., Perkin Trans. 1 1993, 101 and references cited therein. See also:
Sanchez, I. H.; Larraza, M. I.; Rojas, I.; Brena, F. K.; Flores, H. J.
Heterocycles 1985, 23, 3033.

Synthesis of Amaryllidaceae Alkaloids
J. Am. Chem. Soc., Vol. 119, No. 25, 1997 5775
Scheme 1
Scheme 3
protecting group for the C-2 alcohol (montanine numbering)
and found that 15 could be converted cleanly to epoxy ether
18. Epoxy benzyl ether 18 was opened regioselectively with
cyanide ion21 to produce nitrile alcohol 19 in 93% yield.
Alcohol 19 was subsequently silylated to afford intermediate
ether 20 (94% yield)20 having the C-2 and C-3 oxygens
differentially protected.
To continue the synthesis, alkene 20 was hydroborated with
disiamylborane22 to regioselectively generate primary alcohol
21, which then was cleanly oxidized under Swern conditions
to yield aldehyde 22 (Scheme 3). Addition of ethynylmag-
nesium bromide to aldehyde 22 was not particularly stereo-
selective, affording a chromatographically separable 2/1 mixture
of (S)-propargyl alcohol 23 and (R)-alcohol 24, respectively.
In order to establish the configuration of alcohols 23 and 24,
as well as to try to improve the stereoselectivity in this phase
of the synthesis, we investigated the reactions shown in Scheme
4. Using Jones reagent, the mixture of propargyl alcohols 23/
24 could be oxidized to acetylenic ketone 25 (82% yield).
Attempted asymmetric reduction of ketone 25 with (S)-Alpine
borane failed, in all attempts giving a complex product mixture.23
However, enantioselective reduction of ketone 25 using LiAlH4/
Darvon alcohol complex24 provided a 5.3/1 mixture of 24 and
23 in 84% yield. The configuration of the major isomer 24
was tentatively assigned as (R) on the basis of the well-known
propensity of Darvon alcohol to generate this configuration in
LiAlH4 reductions of alkynyl ketones.24 Similarly, reduction
of ketone 25 with LiAlH4/ent-Darvon alcohol25 provided a 1/4.8
mixture of the alcohols 24/23. Although the selectivity here is
somewhat better than in the direct preparation of these propargyl
alcohols from aldehyde 22, the two extra steps involved do not
make this sequence an attractive alternative, particularly since
it was found that both 23 and 24 can be efficiently used for the
total synthesis (Vide infra).
Scheme 2
Thus, (S)-propargyl alcohol 23 could be directly acetylated
to give the desired (S)-acetate 26 in the presence of acetic
anhydride, triethylamine, and a catalytic amount of DMAP in
98% yield (Scheme 5). The (R)-alcohol 24 can also be
converted to the same (S)-acetate 26 by a Mitsunobu inversion
procedure26 using diethyl azodicarboxylate, triphenylphosphine,
acetic acid, and pyridine (86% yield). Thus, we were able to
Results and Discussion
Our synthesis began with known scalemic epoxy alcohol 1519
which was prepared Via Sharpless asymmetric epoxidation of
commercially available divinyl carbinol (Scheme 2). The epoxy
alcohol 15, which could be consistently prepared in 65% yield
on a 25 g scale, showed an ee of >97% as analyzed by 19F
NMR of the corresponding Mosher ester. When epoxy alcohol
15 was subjected to silylation conditions,20 a mixture of epoxy
silyl ether 16 and ring-opened chloro alcohol 17 was produced.
Thus, at this point we chose instead to use a benzyl ether
(21) Behrens, C. H.; Ko, S. Y.; Sharpless, K. B.; Walker, F. J. J. Org.
Chem. 1985, 50, 5687.
(22) (a) Kabalka, G. W.; Hedgecock, H. C., Jr. J. Org. Chem. 1975, 40,
1776. (b) Brown, H. C.; Kulkarni, S. U.; Rao, C. G. Synthesis 1980, 151.
(23) Midland, M. M.; McDowell, D. C.; Hatch, R. L.; Tramontano A. J.
Am. Chem. Soc. 1980, 102, 867.
(24) (a) Brinkmeyer R. S.; Kapoor, V. M. J. Am. Chem. Soc. 1977, 99,
8339. (b) Lopresti, R. J.; Neukom, C.; Saucy, G. J. Org. Chem. 1980, 45,
583.
(25) We thank Professor Larry Overman for a sample of ent-Darvon
alcohol.
(19) (a) Schreiber, S. L.; Schreiber, T. S.; Smith, D. B. J. Am. Chem.
Soc. 1987, 109, 1525. (b) Babine, R. E. Tetrahedron Lett. 1986, 27, 5791.
Distillation of epoxide 15 is recommended. It is necessary, however, to
remove excess peroxide by chromatography prior to distillation.
(20) Wuts, P. G. M.; Bigelow, S. S. J. Org. Chem. 1988, 53, 5023.
(26) Wovkulich, P. M.; Shankaran, K.; Kiegiel, J.; Uskokovic, M. R. J.
Org. Chem. 1993, 58, 832.

5776 J. Am. Chem. Soc., Vol. 119, No. 25, 1997
Jin and Weinreb
Scheme 4
Scheme 7
Scheme 5
We next investigated synthesis of (6-bromopiperonyl)amine
(33). This compound was very easily prepared in 64% yield
by reductive amination28 of known 6-bromopiperonal (32) (eq
2), which was synthesized by bromination of commercially
Scheme 6
available piperonal.29 As delineated in our retrosynthetic plan
in Scheme 1, an allenylsilane imine (12) was to be generated
by direct condensation of an allenylsilane aldehyde 13 and a
6-substituted piperonylamine (14). However, upon treatment
of allenylsilane aldehyde 28 with amine 33 in the presence of
1
4 Å molecular sieves in benzene-d6 at room temperature, H
prepare the requisite (S)-propargyl acetate 26 efficiently, and
we next turned our attention to stereospecific transformation
of this propargyl acetate to the (R)-allenylsilane needed for the
synthesis.
NMR showed that no allenylsilane imine was formed and only
starting materials remained. When this reaction mixture was
heated at 50 °C for 12 h, â-elimination product 34 was formed
rather than the desired allenylsilane imine (eq 3). The same
elimination product was observed when zinc chloride was used
as a catalyst at room temperature.
In 1984, Fleming and Terrett reported that silacupration of
propargyl acetates stereospecifically affords allenylsilanes via
an anti-SN2′ process.27 Fleming’s methodology was applied to
our system, and thus silacupration of (S)-propargyl acetate 26
afforded (R)-allenylsilane 27 in 84% yield. This key transfor-
mation allowed us to stereospecifically transfer the (S)-propargyl
acetate configuration of 26 to the desired (R)-allene configu-
ration of 27, which is crucial to our allenylsilane imino ene
strategy (Vide infra). Reduction of nitrile 27 with diisobutyl-
aluminum hydride (DIBALH) produced allenylsilane aldehyde
28 (78% yield). Using the chemistry described here, scalemic
allenylsilane aldehyde 28 was prepared efficiently in nine steps
and 38% overall yield from chiral epoxy alcohol 15.
One possible explanation for this unexpected result is that
the steric hindrance produced by the o-bromine atom in amine
33 slows formation of the desired imine. Therefore, to test this
possibility, we decided to employ the system lacking this ortho
substituent, with the hope of introducing it at a later stage. In
fact, an allenylsilane imine could be successfully generated upon
treatment of allenylsilane aldehyde 28 with commercially
available piperonylamine (35) (Scheme 7). When this imine
was refluxed in mesitylene for 2 h, we were pleased to find
In order to further test the stereospecificity of the silacupration
step, diastereomeric propargyl acetate 29 was prepared from
(R)-alcohol 24 (Scheme 6). Addition of the Fleming silyl
cuprate to 29 cleanly yielded (S)-allenylsilane nitrile 30 in high
yield. DIBALH reduction of 30 then afforded the diastereo-
meric allenylsilane aldehyde 31.
(28) Ryckman, D. M.; Stevens, R. V. J. Org. Chem. 1987, 52, 4274.
(29) (a) Jung, M. E.; Lam, P. Y.; Mansuri, M. M.; Speltz, L. M. J. Org.
Chem. 1985, 50, 1087. (b) Khanapure, S. P.; Biehl, E. R. J. Org. Chem.
1990, 55, 1471.
(27) Fleming, I.; Terrett, N. K. J. Organomet. Chem. 1984, 264, 99.

Synthesis of Amaryllidaceae Alkaloids
J. Am. Chem. Soc., Vol. 119, No. 25, 1997 5777
Scheme 8
Scheme 9
Scheme 10
that cis-silylacetylene amine 38 was produced as a single
stereoisomeric cycloadduct. This silyl acetylene was im-
mediately desilylated to afford cis-amino alkyne 39 in 66% yield
from aldehyde 28. We believe that product 38 is generated by
a concerted pericyclic imino ene process involving imine
conformations 36 and/or 37 (cf. eq 1). Inspection of models
indicates that both conformations are stereoelectronically capable
of undergoing concerted ene reactions.13 However, it is
important to note that both conformations lead to the same
cycloadduct 38.
1H NMR indicated that the amino and acetylene groups of
39 have the anticipated cis relationship, with a vicinal coupling
constant between H5 and H11a (montanine numbering) of 4.0
Hz (cf. 39a). The coupling constant between H4 and H5 is 10.6
Hz, which indicates that these two hydrogens are trans diaxial.
Although the complete stereochemistry of cycloadduct 39 was
not unambiguously established at this time, we eventually were
able to convert the brominated analog of 39 to the natural
products, thereby confirming the stereochemical assignments
(Vide infra).
In order to further explore the stereospecificity of this imino
ene process, we performed the ene cyclization using diastereo-
meric allenylsilane aldehyde 31. The allenylsilane imine
generated from aldehyde 31 and piperonylamine (35) cyclized
in refluxing mesitylene for 2 h to give a single product (42),
which was then desilylated to cis-amino alkyne 43 (52% yield
over three steps) (Scheme 8). We believe that cycloadduct 42
is generated by a concerted pericyclic imino ene process
involving imine conformations 40 and/or 41. Again, as with
36 and 37, both of these conformations are capable of undergo-
ing concerted ene reactions, and both lead to the identical
cycloadduct 42. One important aspect of these allenylsilane
imino ene reactions which should be reiterated is that the allene
absolute configurations in imines 36/37 and 40/41 control the
relative stereochemistry created between the pair of substituents
at C-2/C-3 Vs the acetylene and amino groups at C-5/C-11a.
As shown in Schemes 7 and 8, the (R)-allene configuration of
allenylsilane imine 36/37 results in amino alkyne 38 which has
these two sets of groups trans to each other, and the (S)-allene
configuration of imine 40/41 results in amino alkyne 42 where
these two pairs have a cis relationship.
original concept of utilizing a piperonylamine analog already
bearing the required C-6 bromine atom.
It is known that aldehydes and ketones can be condensed
with iminophosphoranes to generate imines under mild condi-
tions.30 We thus generated the appropriate iminophosphorane
as shown in Scheme 9. Known piperonyl alcohol 4429b was
converted to iodide 45 and then to azide 46 in high overall yield.
Treatment of this azide with triphenylphosphine produced the
iminophosphorane 47, which without isolation was used for the
next step. Iminophosphorane 47 and allenylsilane aldehyde 28
were heated in mesitylene at 50 °C for 12 h, and then at reflux
for an additional 2 h, to afford a single ene cyclization product
(48), which was immediately desilylated to give amino acetylene
49 (63% based upon aldehyde 28) (Scheme 10). The structure
of this cycloadduct was eventually confirmed by correlation with
the natural alkaloids (Vide infra).
It was felt that it would be best to protect the nitrogen of 49
before continuing the synthetic route. However, the amino
group in this compound appears to be very hindered, and even
under forcing conditions formation of the corresponding N-tosyl
derivative could not be driven to completion.16b Alternatively,
we decided to continue with the free amine, and a Lindlar
hydrogenation was used to convert alkyne 49 to terminal olefin
50 (93%).31 We were pleased to find that exposure of bromo
olefin 50 to Heck cyclization conditions led to seven-membered
At this juncture we turned to an investigation of the
introduction of a bromine at C-6 of the aromatic ring of
cyclization product 39 or a derivative.16b Despite a number of
attempts to effect bromination of various N-protected analogs
of 39, as well as with the olefin derived from Lindlar
hydrogenation of the alkyne functionality, under no circum-
stances could we produce a halogenated substrate needed for
the projected Heck cyclization. We therefore returned to the
(30) See for example: Lambert, P.; Vaultier, M.; Carrie, R. J. Chem.
Soc., Chem. Commun. 1982, 1224.
(31) Amino acetylene 49 could be converted to the N-acetyl derivative
and reduced to the olefin. This intermediate underwent Heck cyclization at
110 °C in 92% yield to the acetamide of 51. However, the acetyl group
was found to be incompatible with the hydroboration step, and thus this
sequence was not pursued.16b

5778 J. Am. Chem. Soc., Vol. 119, No. 25, 1997
Jin and Weinreb
Scheme 11
Scheme 12
exocyclic alkene 51 in good yield. This compound could then
be readily protected as its N-tosyl derivative 52.
With key intermediate 52 in hand, we next turned to studies
on the hydroboration of the exo-methylene group. Disappoint-
ingly, hydroboration of 52 with 1-pyrrolylborane or borane-
THF complex led to 1/1 mixtures of stereoisomeric hydroxy-
methyl compounds 53 and 54 (87% and 67% yields, respec-
tively) (Scheme 11).16b The olefinic substrate was found to be
unreactive toward disiamylborane and 9-BBN, and with Br2BH-
DMS complex, only decomposition occurred.
A search of the literature revealed a possible alternative to
the hydroboration of the olefinic moiety of 52. Danishefsky
and McClure have reported32 an example of an exocyclic alkene
being stereoselectively transformed into a hydroxymethyl group
Via a Lewis acid promoted rearrangement of the derived epoxide.
Although the olefin here was embedded in a very different ring
system from ours, there were enough similarities to warrant
investigating a similar sequence. Therefore, an attempt was first
made to epoxidize alkene 52 with m-chloroperoxybenzoic acid,
but only a complex product mixture was produced. However,
dimethyldioxirane33 produced a 2/1 mixture of epoxide dia-
stereoisomers 55 (Scheme 11). This epoxide mixture was
unstable to silica gel chromatography and was therefore carried
on to the next step without separation. Ring opening of the
mixture of epoxide diastereoisomers 55 induced by ferric
chloride afforded a single rearrangement product aldehyde (57),
which was immediately reduced to the desired alcohol dia-
stereoisomer 54. We believe that the epoxides 55 are opened
by the Lewis acid to first generate benzylic carbocation 56,
which then undergoes a 1,2-hydrogen shift from the least
hindered â-face to afford the kinetic R-aldehyde 57. Immediate
reduction of aldehyde 57 to alcohol 54 without workup proved
crucial, since it was observed that this aldehyde epimerized to
the thermodynamically more stable undesired â-epimer on silica
gel chromatography. The stereochemistry of alcohol 54 was
tentatively assigned as shown on the basis of mechanistic
grounds, but this assumption was soon confirmed by subsequent
correlation with known compounds (Vide infra).
Mitsunobu conditions35 afforded the bridged pentacyclic inter-
mediate 59 (94%). In order to assure that the structure and
stereochemistry assigned to 59 were correct, this intermediate
was desilylated and acetylated to afford the known acetate 60.12
The 1H NMR (500 MHz) spectrum of our enantiomerically pure
acetate 60 was identical to that of a racemic sample provided
by Professor Hoshino.36
At this stage, difficulties arose in what was anticipated to be
a straightforward transformation, i.e., hydrogenolysis of the
O-benzyl group. Thus, hydrogenation of benzyl ether 59 using
palladium on activated carbon in methanol gave no reaction.
Using the more reactive Pearlman’s catalyst (Pd(OH)2 on
carbon) at room temperature, the desired silyl ether alcohol was
produced in only about 5% yield along with unreacted benzyl
ether. Elevating the temperature of the hydrogenolysis led to
a mixture of the desired silyl ether alcohol (21%) and the diol
(71%) resulting from loss of the TBS group. Other debenzyl-
ation methods such as Birch reduction were also tried unsuc-
cessfully. One possible explanation for the problems in the
catalytic hydrogenation is that the steric crowding due to the
adjacent cis-O-benzyl and bulky OTBS groups in 59 slows the
process. In fact, the less crowded acetate benzyl ether 60 could
After the successful transformation of alkene 52 to the desired
hydroxymethyl compound 54, we turned to construction of the
bridged ring system of the montanine alkaloids. Thus, tolu-
enesulfonamide 54 was deprotected by treatment with sodium
naphthalenide34 at -78 °C to generate secondary amine 58 in
96% yield (Scheme 12). Cyclodehydration of amine 58 under
(32) McClure, K. F.; Danishefsky, S. J. J. Am. Chem. Soc. 1993, 115,
6094.
(33) (a) Murray, R. W.; Jeyaraman, R. J. Org. Chem. 1985, 50, 2847.
(b) Adam, W.; Bialas, J.; Hadjiarapoglou, L. Chem. Ber. 1991, 124, 2377.
(34) McIntosh, J. M.; Matassa, L. C. J. Org. Chem. 1988, 53, 4452.
(35) (a) Tsunoda, T.; Ozaki, F.; Shirakata, N.; Tamaoka, Y.; Yamamoto,
H.; Ito, S. Tetrahedron Lett. 1996, 37, 2463. (b) Goldstein, D. M.; Wipf,
P. Tetrahedron Lett. 1996, 37, 739.
(36) We are grateful to Professor O. Hoshino for copies of the NMR
spectra of intermediate 60, as well as racemic montanine and coccinine.

Synthesis of Amaryllidaceae Alkaloids
J. Am. Chem. Soc., Vol. 119, No. 25, 1997 5779
Scheme 13
Scheme 15
Scheme 14
O-TBS) suffered from low yields. In the best case, the enone
could be generated in only 26% yield by oxidation of the
saturated ketone with DDQ. It was therefore necessary for us
to find a suitable alternative method for enone formation from
saturated ketone 65.
We expected that regioselective conversion of ketone 65 to
a kinetic silyl enol ether, followed by oxidation, should be
feasible. Therefore, ketone 65 was first cleanly converted to
the silyl enol ether 66 following the procedure developed by
Corey and Gross.40 Although oxidation of silyl enol ether 66
to R,â-unsaturated ketone 67 with DDQ41 was unsuccessful,
we were pleased to find that 66 was converted to the desired
enone by the Saegusa method (Pd(OAc)2 in acetonitrile)42 in
67% isolated yield (81% yield based on recovered starting
ketone 65) for the two steps.
be hydrogenated using Pearlman’s catalyst in methanol at room
temperature to produce the corresponding alcohol acetate 61 in
94% yield. Although alcohol 61 might potentially be used in
the synthesis, the extra deprotection and reprotection steps
required did not make this sequence very attractive. Since it
also seemed possible that the basic nitrogen in 59 was also
contributing to the slow rate of benzyl group hydrogenolysis,37
it was decided to return to N-protected tetracyclic sulfonamide
54.
Indeed catalytic hydrogenation of N-tosyl compound 54 using
Pd/C in methanol at room temperature produced the desired
alcohol 62 in 97% yield (Scheme 13). Deprotection of
toluenesulfonamide 62 then afforded amino diol 63 (96% yield).
Cyclodehydration of 63 under standard Mitsunobu conditions
did generate the desired bridged intermediate 64, but in only
42% yield. We were pleased to find, however, that treatment
of 63 with imidazole, triphenylphosphine, and iodine38 at 0 °C
for 10 min produced 64 in 82% overall yield for the three steps
from 54.
With pentacyclic alcohol 64 now in hand, the stage was set
for introduction of the C-1/C-11a double bond and effecting
the final functional group tranformations to produce the natural
products. Thus, alcohol 64 was cleanly oxidized to ketone 65
with N-methylmorpholine N-oxide and tetrapropylammonium
perruthenate39 in the presence of 4 Å molecular sieves (Scheme
14). In Hoshino’s synthesis,12 all attempts to form an enone
from a ketone very closely related to 65 (O-benzyl rather than
Conversion of enone 67 to the 5,11-methanomorphanthridine
alkaloids 1-3 proved to be straightforward. Desilylation of
67 with tetrabutylammonium fluoride produced the known
R-hydroxy enone 6811 in 98% yield. 1H NMR, 13C NMR, and
low- and high-resolution mass spectra of our synthetic 68 were
identical to those previously reported by Overman and Shim.11,43
The optical rotation [[R]23 ) -55.9° (c ) 0.272, MeOH),
546
[R]23D ) -50.7° (c ) 0.272, MeOH)] of our synthetic material
was in good agreement with that reported [[R]25 ) -47.7°
546
(c ) 0.10, MeOH)].11 Overman and Shim have found that
enantiomerically pure 68 could be reduced to (-)-pancracine
(3) with sodium triacetoxyborohydride in 62% yield.11
We could also prepare (-)-coccinine (2) from 68 by the
chemistry shown in Scheme 15. Dimethyl ketal 69 was
synthesized in 91% yield by treatment of enone 68 with
p-toluenesulfonic acid and trimethyl orthoformate in methanol.
When 69 was treated with DIBALH in toluene at room
temperature,12 it underwent both ketone reduction and cleavage
of the TBS group to afford (-)-coccinine (2) in 81% yield. (-)-
Montanine (1) was also produced in this reaction as a minor
product (8%). We believe that the large OTBS group in
intermediate 70 shields the â face of the molecule, therefore
leading to (-)-coccinine (2) as the major diastereoisomer of
the reduction. The formation of a small amount of (-)-
montanine (1) is possibly due to some silyl group cleavage prior
1
to hydride reduction of the ketal (Vide infra). The H NMR
(37) Cf. Czech, B. P.; Bartsch, R. A. J. Org. Chem. 1984, 49, 4076.
(38) (a) Garegg, P. J.; Samuelsson, B. J. Chem. Soc., Chem. Commun.
1979, 978. (b) Singh, A. K.; Bakshi, R. K.; Corey, E. J. J. Am. Chem. Soc.
1987, 109, 6187.
(39) Griffith, W. P.; Ley, S. V.; Whitcombe, G. P.; White, A. D. J. Chem.
Soc., Chem. Commun. 1987, 1625.
(40) Corey, E. J.; Gross, A. W. Tetrahedron Lett. 1984, 25, 495.
(41) Fleming, I.; Paterson, I. Synthesis 1979, 736.
(42) Ito, Y.; Hirao, T.; Saegusa, T. J. Org. Chem. 1978, 43, 1011.
(43) We thank Professor Larry Overman for copies of the 1H and 13C
NMR spectra of intermediate 68 and a sample of (-)-pancracine.

5780 J. Am. Chem. Soc., Vol. 119, No. 25, 1997
Jin and Weinreb
and mesitylene were distilled from CaH2, and methanol was distilled
from magnesium turnings.
spectrum of our synthetic (-)-coccinine (2) was identical to
that of Hoshino’s racemic material.12,36 The optical rotation
[[R]25 ) -161° (c ) 0.101, EtOH)] of our synthetic 2 was
Preparation of Benzyl Ether Epoxide 18. To a solution of 5.00 g
(50.0 mmol) of scalemic epoxide 15 (>97% ee),19 7.1 mL (60.0 mmol)
of benzyl bromide, and 1.85 g (5.00 mmol) of tetrabutylammonium
iodide in 125 mL of dry THF at -20 °C under argon was added 1.39
g (55.0 mmol, 95%) of NaH. The mixture was allowed to warm to
room temperature (rt) over 2 h. After the mixture was stirred at rt for
another 2 h, 80 mL of saturated aqueous NH4Cl solution was added at
0 °C. The mixture was stirred for 10 min and extracted three times
with 60 mL portions of ether. The combined organic layers were
washed with brine, dried over MgSO4, and concentrated in Vacuo. The
residue was purified by flash chromatography on silica gel, eluting with
D
consistent with that reported for natural (-)-coccinine [[R]27
D
) -188.8° (c ) 1.9, EtOH)].3 The 13C NMR spectrum and
low- and high-resolution mass spectra of our synthetic 2 also
confirmed the structural assignment.
Interestingly, when R-hydroxy ketal 71, generated from
desilylation of ketal 69, was treated with DIBALH in toluene
(Scheme 15), (-)-montanine (1) (41% isolated, 47% based on
recovered starting material) along with (-)-coccinine (2) (39%
isolated, 44% based on recovered starting material) was formed.
Therefore, without the bulky silyl ether group R to the ketal,
the diastereoselectivity of the DIBALH reduction drops sharply.
The 1H NMR spectrum of our synthetic (-)-montanine (1) was
identical to that obtained by Hoshino and co-workers.12,36 The
ethyl acetate/hexanes (1/19), to produce benzyl ether 18 as a clear oil
1
(9.12 g, 96%): [R]20 ) -12.2° (c ) 0.97, benzene); H NMR (300
D
MHz, CDCl3) δ 2.69 (dd, J ) 2.6, 5.3 Hz, 1 H), 2.78 (dd, J ) 4.0, 5.3
Hz, 1 H), 3.08-3.12 (m, 1 H), 3.82-3.87 (m, 1 H), 4.46 (d, J ) 12.1
Hz, 1 H), 4.64 (d, J ) 12.1 Hz, 1 H), 5.36-5.38 (m, 1 H), 5.40-5.42
(m, 1 H), 5.86 (ddd, J ) 7.3, 10.9, 16.7 Hz, 1 H), 7.27 -7.37 (m, 5
H); 13C NMR (75 MHz, CDCl3) δ 44.5, 53.0, 70.4, 79.1, 119.3, 127.4,
127.5 (2 C), 128.2 (2 C), 134.3, 138.0; IR (film) 3080-2940, 2930-
2780 cm-1; CIMS m/z (relative intensity) 190 (M+ + H, 0.5), 189 (M+,
0.7), 173 (1), 107 (11), 91 (100); HRMS calcd for C12H14O2 190.0994,
found 190.1001.
optical rotation [[R]25 ) -83° (c ) 0.06, CHCl3)] of our
D
synthetic (-)-1 corresponded with that reported for natural (-)-
montanine [[R]26D ) -87.6° (c ) 0.6, CHCl3)].3 The identity
of our synthetic (-)-1 was also confirmed by its 13C NMR
spectrum, HMBC NMR, and low- and high-resolution mass
spectra.
Preparation of Nitrile 19. To a solution of 1.80 g (9.50 mmol) of
epoxide 18 in 40 mL of MeOH under argon was added 0.93 g (14.3
mmol) of KCN. The mixture was gently refluxed at 70 °C for 2.5 h.
After the mixture was cooled to rt, 40 mL of water was added. The
aqueous layer was extracted three times with 80 mL portions of ether.
The combined organic layers were washed with brine, dried over
MgSO4, and concentrated in Vacuo. The residue was purified by flash
chromatography on silica gel, eluting with ethyl acetate/hexanes (1/2),
In addition, alcohol 54 was debenzylated and desilylated to
the known triol 7212 in 97% yield (eq 4). Hoshino and co-
to produce nitrile 19 as a yellow oil (1.92 g, 93%): [R]20 ) -57.7°
D
1
(c ) 1.03, benzene); H NMR (300 MHz, CDCl3) δ 2.56 (d, J ) 5.8
Hz, 2 H), 3.35 (br d, J ) 5.3 Hz, 1 H), 3.79-3.92 (m, 2 H), 4.39 (d,
J ) 11.5 Hz, 1 H), 4.64 (d, J ) 11.5 Hz, 1 H), 5.40-5.50 (m, 2 H),
5.79 (ddd, J ) 7.5, 10.5, 17.1 Hz, 1 H), 7.14-7.43 (m, 5 H); 13C NMR
(75 MHz, CDCl3) δ 21.4, 68.8, 70.4, 81.8, 117.7, 121.2, 127.58 (2 C),
workers have previously converted triol 72 to racemic brunsvig-
ine (4) in several steps.12 Therefore, we have also completed a
formal enantioselective total synthesis of (-)-brunsvigine (4).
127.64, 128.2 (2 C), 133.7, 137.3; IR (film) 3600-3200, 2750 cm-1
;
CIMS m/z (relative intensity) 218 (M+ + H, 78), 91 (100); HRMS
calcd for C13H15NO2 217.1103, found 217.1117.
Conclusion
We have developed a new type of thermal intramolecular
concerted ene reaction of allenylsilane imines which has been
successfully used in total syntheses of the 5,11-methanomor-
phanthridine Amaryllidaceae alkaloids (-)-montanine (1), (-)-
coccinine (2), and (-)-pancracine (3). These complex penta-
cyclic natural products were synthesized from readily available
enantiomerically pure epoxy alcohol 15 in about 25 steps. The
key features of the synthetic strategy include (1) a stereospecific
thermal imino ene cyclization of allenylsilane imine derived
from aldehyde 28 to afford key intermediate amino alkyne 48,
(2) an intramolecular Heck reaction of bromo alkene 50 to
produce a seven-membered ring containing tetracycle 51, and
(3) stereospecific formation of hydroxymethyl compound 54
Via epoxidation of the alkene 52, followed by a Lewis acid
catalyzed epoxide ring opening/rearrangement. We plan to use
this and related methodology in approaches to other natural
product targets.
Preparation of Silyl Ether 20. To a solution of 5.20 g (24.0 mmol)
of alcohol 19 in 120 mL of DMF at 0 °C under argon was added 6.53
g (96.0 mmol) of imidazole, followed by 7.22 g (48.0 mmol) of tert-
butyldimethylsilyl chloride. After the mixture was stirred for 16 h at
rt, 100 mL of water was added. The mixture was extracted three times
with 100 mL portions of ether. The combined organic layers were
washed with brine, dried over MgSO4, and concentrated in Vacuo. The
residue was purified by flash chromatography on silica gel, eluting with
ethyl acetate/hexanes (3/97), to produce silyl ether 20 as a clear oil
1
(7.47 g, 94%): [R]20 ) -25.7° (c ) 1.21, benzene); H NMR (300
D
MHz, CDCl3) δ 0.12 (s, 3 H), 0.18 (s, 3 H), 0.96 (s, 9 H), 2.53 (dd, J
) 4.3, 16.8 Hz, 1 H), 2.74 (dd, J ) 5.6, 16.8 Hz, 1 H), 3.83-3.87 (m,
1 H), 3.93-3.97 (m, 1 H), 4.45 (d, J ) 11.5 Hz, 1 H), 4.66 (d, J )
11.5 Hz, 1 H), 5.39-5.47 (m, 2 H), 5.79 (ddd, J ) 7.2, 11.0, 18.1 Hz,
1 H), 7.27-7.39 (m, 5 H); 13C NMR (75 MHz, CDCl3) δ -4.9, -4.5,
17.8, 22.8, 25.6 (3 C), 70.6, 70.7, 82.3, 117.6, 120.3, 127.6, 127.7 (2
C), 128.2 (2 C), 134.8, 137.7; IR (film) 3100-2840, 2250 cm-1; CIMS
m/z (relative intensity) 332 (M+ + H, 33), 91 (100); HRMS calcd for
C15H20NO2Si (M+ - t-Bu) 274.1263, found 274.1254.
Preparation of Alcohol 21. To 44.2 mL (44.2 mmol) of borane-
THF complex (1.0 M solution in THF) under argon at 0 °C was added
dropwise 9.84 mL (88.4 mmol) of 2-methyl-2-butene. After being
stirred for 1 h at 0 °C, the mixture was added dropwise through a
cannula to a solution of 12.19 g (36.8 mmol) of alkene 20 in 75 mL of
THF at 0 °C under argon. The mixture was stirred for 6 h at rt. After
the mixture was cooled to -20 °C, 23.7 mL of H2O2 (30% in H2O)
and 77.3 mL of aqueous 3 M NaOH solution were added slowly. The
resulting mixture was stirred at rt for 10 min and extracted three times
with 80 mL portions of ether. The combined organic layers were dried
over Na2SO4 and concentrated in Vacuo. The residue was purified by
flash chromatography on silica gel, eluting with ethyl acetate/hexanes
Experimental Section
General Methods. Low-resolution mass spectra (MS) were obtained
at 50-70 eV by electron impact (EI). Chemical ionization mass spectra
(CIMS) were obtained using isobutane as a carrier gas. Optical rotations
were obtained at ambient temperature. Flash chromatography was
performed using EM Science silica gel 60. Analytical and preparative
TLC were performed on EM silica gel 60 PF254. HPLC was done using
a Beckman Ultrasphere SI 5 mm, 10.0 mm × 25 cm column. All
chemicals were purchased from Aldrich Chemical Co. unless otherwise
noted. THF, benzene, and ether were dried over and distilled from
sodium/benzophenone ketyl. CH2Cl2, toluene, NEt3, DMF, pyridine,

Synthesis of Amaryllidaceae Alkaloids
J. Am. Chem. Soc., Vol. 119, No. 25, 1997 5781
(1/2), to produce alcohol 21 as a clear oil (11.28 g, 88%): [R]20
)
NaHCO3 solution was added, and stirring was continued for 15 min.
The aqueous layer was extracted three times with 15 mL portions of
CH2Cl2. The combined organic layers were washed with brine, dried
over MgSO4, and concentrated in Vacuo. The residue was purified by
flash chromatography on silica gel, eluting with ethyl acetate/hexanes
D
1
+4.9° (c ) 1.08, benzene); H NMR (300 MHz, CDCl3) δ 0.14 (s, 3
H), 0.18 (s, 3 H), 0.95 (s, 9 H), 1.69-1.82 (m, 2 H), 2.21 (br s, 1 H),
2.52 (dd, J ) 4.5, 16.8 Hz, 1 H), 2.70 (dd, J ) 6.2, 16.8 Hz, 1 H),
3.71-3.75 (m, 3 H), 4.01-4.08 (m, 1 H), 4.62 (d, J ) 11.2 Hz, 1 H),
4.78 (d, J ) 11.2 Hz, 1 H), 7.27-7.40 (m, 5 H); 13C NMR (75 Mz,
CDCl3) δ -4.9, -4.6, 17.8, 22.2, 25.6 (3 C), 33.0, 59.3, 70.9, 73.4,
79.7, 117.9, 127.9 (3 C), 128.4 (2 C), 137.7; IR (film) 3600-3300,
2250 cm-1; CIMS m/z (relative intensity) 350 (M+ + H, 92), 91 (100);
HRMS calcd for C19H31NO3Si 349.2073, found 349.2060.
Preparation of Aldehyde 22. To a solution of 13.7 mL (27.4 mmol)
of oxalyl chloride (2.0 M solution in CH2Cl2) in 30 mL of dry CH2Cl2
at -78 °C under argon was slowly added a solution of 4.30 mL (60.4
mmol) of DMSO in 12 mL of dry CH2Cl2 while the mixture was kept
below -65 °C. After the mixture was stirred for 10 min at -78 °C,
6.39 g (18.3 mmol) of alcohol 21 in 10 mL of CH2Cl2 was added below
-65 °C. The mixture was stirred at -78 °C for 30 min and
triethylamine (15.3 mL (109.9 mmol)) was added at -78 °C. The
mixture was allowed to warm to rt, and 30 mL of water was added.
The aqueous layer was extracted three times with 50 mL portions of
CH2Cl2. The combined organic layers were washed with aqueous 5%
HCl solution, and brine, dried over MgSO4, and concentrated in Vacuo.
The residue was purified by flash chromatography on silica gel, eluting
(1/5), to produce propargyl acetate 26 as a clear oil (714 mg, 98%):
1
[R]20 ) -1.7° (c ) 1.27, benzene); H NMR (300 MHz, CDCl3) δ
D
0.14 (s, 3 H), 0.18 (s, 3 H), 0.96 (s, 9 H), 1.88-2.02 (m, 2 H), 2.04 (s,
3 H), 2.46 (dd, J ) 4.4, 16.9 Hz, 1 H), 2.54 (d, J ) 2.0 Hz, 1 H), 2.65
(dd, J ) 6.6, 16.9 Hz, 1 H), 3.76-3.82 (m, 1 H), 4.02-4.06 (m, 1 H),
4.60 (d, J ) 11.2 Hz, 1 H), 4.77 (d, J ) 11.2 Hz, 1 H), 5.50 (ddd, J
) 2.0, 6.4, 8.5 Hz, 1 H), 7.27-7.36 (m, 5 H); 13C NMR (75 MHz,
CDCl3) δ -5.1, -4.7, 17.7, 20.6, 21.9, 25.5 (3 C), 36.2, 61.4, 71.0,
73.5, 74.5, 78.5, 80.3, 117.6, 127.7 (3 C), 128.2 (2 C), 137.4, 169.1;
IR (film) 3260, 2240, 2105, 1740 cm-1; CIMS m/z (relative intensity)
416 (M+ + H, 100); HRMS calcd for C23H33NO4Si 415.2179, found
415.2158.
Preparation of (S)-Propargyl Acetate 26 from (R)-Propargyl
Alcohol 24. To a solution of 280 mg (0.751 mmol) of propargyl alcohol
24, 788 mg (3.00 mmol) of Ph3P, 0.12 mL (1.50 mmol) of pyridine,
and 0.22 mL (3.75 mmol) of acetic acid in 10 mL of THF was added
a solution of 523 mg (3.00 mmol) of DEAD in 2 mL of THF at -45
°C under argon. After being stirred at 0 °C for 16 h, the mixture was
taken up in 50 mL of ether and washed with saturated aqueous NaHCO3,
5% aqueous HCl solution, and brine. The organic layer was dried over
MgSO4 and concentrated in Vacuo. The residue was purified by flash
chromatography on silica gel, eluting with ethyl acetate/hexanes (1/5),
to produce propargyl acetate 26 as a clear oil (266 mg, 86%).
with ethyl acetate/hexanes (1/5), to produce aldehyde 22 as a clear oil
1
(6.10 g, 96%): [R]20 ) -4.7° (c ) 1.12, benzene); H NMR (300
D
MHz, CDCl3) δ 0.14 (s, 3 H), 0.17 (s, 3 H), 0.95 (s, 9 H), 2.50 (dd, J
) 4.3, 16.8 Hz, 1 H), 2.693 (dd, J ) 5.0, 16.8 Hz, 1 H), 2.695 (dd, J
) 1.9, 5.4 Hz, 2 H), 3.99-4.07 (m, 2 H), 4.63 (d, J ) 11.2 Hz, 1 H),
4.68 (d, J ) 11.2 Hz, 1 H), 7.30-7.37 (m, 5 H), 9.76 (t, J ) 1.9 Hz,
1 H); 13C NMR (75 MHz, CDCl3) δ -5.0, -4.8, 17.7, 22.5, 25.4 (3
C), 44.6, 70.0, 72.8, 76.5, 117.2, 127.7 (2 C), 127.8, 128.3 (2 C), 137.2,
199.7; IR (film) 2980-2800, 2250, 1725 cm-1; CIMS m/z (relative
Preparation of Allene Nitrile 27. To a mixture of 427 mg (61.5
mmol) of lithium shot in 12 mL of dry THF under argon at 0 °C was
added dropwise a solution of 2.103 g (12.3 mmol) of dimethylphenyl-
silyl chloride in 2 mL of dry THF, and the mixture was warmed to rt.
After being stirred for 12 h, the red solution was added to a suspension
of CuCN (552 mg, 6.15 mmol) in 80 mL of dry THF under argon at
0 °C, and the mixture was stirred for 30 min. A solution of 2.556 g
(6.15 mmol) of propargyl acetate 26 in 10 mL of dry THF was added
to the mixture at -96 °C by syringe pump over 15 min. After being
stirred for 4 h at -90 °C, the mixture was poured into 100 mL of
saturated aqueous NH4Cl solution. The mixture was stirred for 1 h
and extracted three times with 100 mL portions of ether. The combined
organic layers were washed with brine, dried over MgSO4, and
concentrated in Vacuo. The residue was purified by flash chromatog-
raphy on silica gel, eluting with ethyl acetate/hexanes (1/19), to produce
intensity) 348 (M+ + H, 100); HRMS calcd for C18H26NO3Si (M+
Me) 332.1682, found 332.1675.
-
Preparation of Propargyl Alcohols 23 and 24. To a solution of
5.45 g (15.71 mmol) of aldehyde 22 in 80 mL of dry CH2Cl2 under
argon at -78 °C was added dropwise 34.6 mL (17.30 mmol) of
ethynylmagnesium bromide (0.5 M solution in THF), and the mixture
was slowly warmed to 0 °C. After the mixture was stirred for 12 h at
0 °C, 50 mL of 5% aqueous HCl solution was added at 0 °C. The
mixture was warmed to rt and stirred for 10 min. The aqueous layer
was extracted three times with 50 mL portions of ether. The combined
organic layers were washed with brine, dried over MgSO4, and
concentrated in Vacuo. The residue was purified by flash chromatog-
raphy on silica gel, eluting with CH2Cl2, to produce 23 and 24 as clear
oils. 23 (3.47 g, 59%): [R]20D ) +10.0° (c ) 1.10, benzene); 1H NMR
(300 MHz, CDCl3) δ 0.16 (s, 3 H), 0.19 (s, 3 H), 0.97 (s, 9 H), 1.79-
2.05 (m, 2 H), 2.50 (dd, J ) 4.6, 16.8 Hz, 1 H), 2.51 (d, J ) 2.0 Hz,
1 H), 2.68 (dd, J ) 6.7, 16.8 Hz, 1 H), 3.78-3.84 (m, 1 H), 4.04-
4.10 (m, 1 H), 4.52-4.58 (m, 1 H), 4.59 (d, J ) 11.2 Hz, 1 H), 4.79
(d, J ) 11.2 Hz, 1 H), 7.30-7.36 (m, 5 H); 13C NMR (75 MHz, CDCl3)
δ -5.1, -4.6, 17.8, 21.9, 25.6 (3 C), 38.9, 59.6, 71.1, 73.5 (2 C), 79.5,
84.1, 117.9, 127.78, 127.83 (2 C), 128.3 (2 C), 137.6; IR (film) 3600-
3200, 2990-2800, 2250, 2110 cm-1; CIMS m/z (relative intensity) 374
(M+ + H, 90), 91 (100), HRMS calcd for C21H31NO3Si 373.2073, found
allene 27 as a clear oil (2.54 g, 84%): [R]20 ) -29.1° (c ) 1.12,
D
1
benzene); H NMR (200 MHz, CDCl3) δ 0.11 (s, 3 H), 0.16 (s, 3 H),
0.37 (s, 6 H), 0.94 (s, 9 H), 2.28-2.40 (m, 2 H), 2.49 (dd, J ) 4.2,
16.8 Hz, 1 H), 2.76 (dd, J ) 5.3, 16.8 Hz, 1 H), 3.51-3.56 (m, 1 H),
3.94-3.99 (m, 1 H), 4.62 (s, 2 H), 4.81-4.91 (m, 1 H), 5.07-5.12
(m, 1 H), 7.30 -7.59 (m, 10 H); 13C NMR (75 MHz, CDCl3) δ -4.6
(2 C), -2.3 (2 C), 17.9, 22.6, 25.7 (3 C), 29.2, 69.9, 72.9, 78.8, 81.1,
81.3, 117.9, 127.8 (4 C), 128.4 (2 C), 129.2, 133.6 (2 C), 133.9, 138.0,
138.2, 211.5; IR (film) 3080-2840, 2240, 1930 cm-1; EIMS m/z
(relative intensity) 491 (M+, 3), 434 (8), 81 (100); HRMS calcd for
C29H41NO2Si2 491.2676, found 491.2675.
Preparation of Allene Aldehyde 28. To a solution of 1.705 g (3.47
mmol) of allene nitrile 27 in 35 mL of dry toluene under argon at -78
°C was added 6.25 mL (6.25 mmol) of DIBALH (1.0 M solution in
hexanes). After the mixture was slowly warmed to 0 °C over 2 h, 20
mL of aqueous 5% HCl solution was added. The mixture was stirred
for 20 min, and was extracted three times with 50 mL portions of ether.
The combined organic layers were washed with brine, dried over
MgSO4, and concentrated in Vacuo. The residue was purified by flash
chromatography on silica gel, eluting with ethyl acetate/hexanes (2/
373.2057. 24 (1.76 g, 30%): [R]20 ) +23.5° (c ) 1.30, benzene);
D
1H NMR (300 MHz, CDCl3) δ 0.16 (s, 3 H), 0.20 (s, 3 H), 0.97 (s, 9
H), 1.80-1.93 (m, 2 H), 2.50 (dd, J ) 4.5, 16.8 Hz, 1 H), 2.51 (d, J
) 2.1 Hz, 1 H), 2.66 (dd, J ) 6.6, 16.8 Hz, 1 H), 3.96-4.00 (m, 1 H),
4.02-4.07 (m, 1 H), 4.51-4.57 (m, 1 H), 4.64 (d, J ) 10.9 Hz, 1 H),
4.84 (d, J ) 10.9 Hz, 1 H), 7.29-7.38 (m, 5 H); 13C NMR (75 MHz,
CDCl3) δ -5.0, -4.6, 17.8, 21.9, 25.6 (3 C), 37.9, 59.0, 71.0, 73.1,
73.8, 79.1, 84.4, 117.9, 127.9, 128.0 (2 C), 128.4 (2 C), 137.4; IR (film)
3600-3200, 2990-2800, 2250, 2110 cm-1; CIMS m/z (relative
intensity) 374 (M+ + H, 90), 91 (100); HRMS calcd for C21H31NO3Si
373.2073, found 373.2057.
98), to produce allene aldehyde 28 as a clear oil (1.33 g, 78%): [R]20
D
) -20.9° (c ) 0.98, benzene); 1H NMR (200 MHz, CDCl3) δ 0.07 (s,
3 H), 0.09 (s, 3 H), 0.35 (s, 6 H), 0.88 (s, 9 H), 2.20-2.36 (m, 2 H),
2.50 (ddd, J ) 2.4, 5.4, 16.1 Hz, 1 H), 2.69 (ddd, J ) 2.5, 5.6, 16.1
Hz, 1 H), 3.39-3.47 (m, 1 H), 4.23-4.31 (m, 1 H), 4.56 (d, J ) 11.5
Hz, 1 H), 4.64 (d, J ) 11.5 Hz, 1 H), 4.82-4.93 (m, 1 H), 5.06-5.12
(m, 1 H), 7.26-7.58 (m, 10 H), 9.81 (dd, J ) 2.4, 2.5 Hz, 1 H); 13C
NMR (75 MHz, CDCl3) δ -4.7, -4.5, -2.2 (2 C), 17.9, 25.8 (3 C),
29.9, 47.3, 70.3, 72.8, 79.7, 81.1, 82.8, 127.6, 127.7 (2 C), 127.8 (3
Preparation of (S)-Propargyl Acetate 26 from (S)-Propargyl
Alcohol 23. To a solution of 656 mg (1.76 mmol) of propargyl alcohol
23 in 15 mL of CH2Cl2 at 0 °C under argon were added 0.49 mL (3.52
mmol) of triethylamine and a catalytic amount of DMAP, followed by
the dropwise addition of 0.25 mL (2.64 mmol) of acetic anhydride.
After the mixture was stirred for 12 h at rt, 10 mL of saturated aqueous

5782 J. Am. Chem. Soc., Vol. 119, No. 25, 1997
Jin and Weinreb
C), 128.3 (2 C), 129.1, 133.6 (2 C), 138.2, 201.3, 211.4; IR (film)
3080-2840, 2700, 1930, 1720 cm-1; CIMS m/z (relative intensity) 495
(M+ + H, 2), 477 (4), 437 (3), 91 (100); HRMS calcd for C29H42O3Si2
494.2672, found 494.2689.
solution, and brine. The organic layer was dried over MgSO4 and
concentrated in Vacuo to produce iodide 45 as a white solid suitable
for use in the next step without further purification (381 mg, 94%):
1H NMR (200 MHz, CDCl3) δ 4.48 (s, 2 H), 5.97 (s, 2 H), 6.88 (s, 1
H), 6.96 (s, 1 H).
Preparation of (R)-Propargyl Acetate 29. Following the procedure
for preparation of (S)-propargyl acetate 26 from (S)-propargyl alcohol
23, (R)-propargyl alcohol 24 (196 mg, 0.525 mmol) was converted to
(R)-propargyl acetate 29 (206 mg, 94%): 1H NMR (200 MHz, CDCl3)
δ 0.10 (s, 3 H), 0.13 (s, 3 H), 0.91 (s, 9 H), 1.94-2.08 (m, 2 H), 1.98
(s, 3 H), 2.45 (dd, J ) 4.9, 17.0 Hz, 1 H), 2.46 (d, J ) 2.2 Hz, 1 H),
2.68 (dd, J ) 6.3, 17.0 Hz, 1 H), 3.59-3.67 (m, 1 H), 3.97-4.08 (m,
1 H), 4.50 (d, J ) 11.1 Hz, 1 H), 4.71 (d, J ) 11.1 Hz, 1 H), 5.41-
5.51 (m, 1 H), 7.28-7.37 (m, 5 H); IR (film) 3260, 2240, 2105, 1740
cm-1; CIMS m/z (relative intensity) 416 (M+ + H, 100); HRMS calcd
for C23H33NO4Si 415.2179, found 415.2158.
Preparation of Allene Nitrile 30. Following the procedure for
preparation of allene 27, propargyl acetate 29 (402 mg, 0.969 mmol)
was converted to allene 30 (408 mg, 86%): 1H NMR (300 MHz,
CDCl3) δ 0.11 (s, 3 H), 0.16 (s, 3 H), 0.37 (s, 6 H), 0.94 (s, 9 H),
2.15-2.46 (m, 2 H), 2.49 (dd, J ) 4.1, 16.8 Hz, 1 H), 2.74 (dd, J )
5.9, 16.8 Hz, 1 H), 3.56 (q, J ) 5.6 Hz, 1 H), 3.92-4.00 (m, 1 H),
4.65 (s, 2 H), 4.80-4.90 (m, 1 H), 5.10-5.16 (m, 1 H), 7.30 -7.57
(m, 10 H); IR (film) 3080-2840, 2240, 1930 cm-1; EIMS m/z (relative
intensity) 491 (M+, 3), 434 (8), 81 (100); HRMS calcd for C29H41-
NO2Si2 491.2676, found 491.2675.
Preparation of Allene Aldehyde 31. Following the procedure for
preparation of aldehyde 28, nitrile 30 (408 mg, 0.831 mmol) was
converted to aldehyde 31 (303 mg, 74%): 1H NMR (200 MHz, CDCl3)
δ 0.03 (s, 3 H), 0.05 (s, 3 H), 0.31 (s, 6 H), 0.85 (s, 9 H), 2.16-2.26
(m, 2 H), 2.48 (ddd, J ) 2.3, 5.1, 16.2 Hz, 1 H), 2.68 (ddd, J ) 2.3,
5.9, 16.2 Hz, 1 H), 3.43-3.52 (m, 1 H), 4.23-4.33 (m, 1 H), 4.59 (d,
J ) 11.6 Hz, 1 H), 4.69 (d, J ) 11.6 Hz, 1 H), 4.85 (q, J ) 7.3 Hz,
1 H), 5.04-5.12 (m, 1 H), 7.29-7.58 (m, 10 H), 9.84 (t, J ) 2.3 Hz,
1 H); IR (film) 3080-2840, 2700, 1930, 1720 cm-1; CIMS m/z (relative
intensity) 495 (M+ + H, 2), 477 (4), 437 (3), 91 (100); HRMS calcd
for C29H42O3Si2 494.2672, found 494.2689.
Preparation of Cycloadduct 39. To a solution of 580 mg (1.17
mmol) of aldehyde 28 in 20 mL of mesitylene was added a solution of
186 mg (1.23 mmol) of piperonylamine (35) in 3 mL of mesitylene at
rt under argon. After being stirred at rt for 1 h, the mixture was refluxed
for 2 h. The mixture was cooled to 0 °C, and 20 mL of THF and 1.17
mL (1.17 mmol) of tetrabutylammonium fluoride (1.0 M solution in
THF) were added. After the mixture was stirred at 0 °C for 1 h, 20
mL of water was added. The mixture was extracted three times with
30 mL portions of ether. The combined organic layers were washed
with brine, dried over K2CO3. and concentrated in Vacuo. The residue
was purified by flash chromatography on silica gel, eluting with ethyl
acetate/hexanes (1/19), to produce amine 39 as a yellow oil (378 mg,
66% from aldehyde 28): 1H NMR (200 MHz, CDCl3) δ 0.03 (s, 6 H),
0.83 (s, 9 H), 1.45-1.63 (m, 2 H), 1.75-1.97 (m, 2 H), 2.06 (d, J )
2.5 Hz, 1 H), 2.97-3.04 (m, 1 H), 3.05-3.15 (m, 1 H), 3.58-3.68
(m, 1 H), 3.62 (d, J ) 12.9 Hz, 1 H), 3.73 (d, J ) 12.9 Hz, 1 H),
4.12-4.18 (m, 1 H), 4.45-4.62 (m, 2 H), 5.92 (s, 2 H), 6.71-6.79
(m, 2 H), 6.83 (s, 1 H), 7.25-7.35 (m, 5 H); IR (film) 3300, 2100
cm-1; EIMS m/z (relative intensity) 493 (M+, 2), 436 (7); HRMS calcd
for C29H39NO4Si 493.2648, found 493.2625.
To a solution of 0.810 g (2.38 mmol) of iodide 45 in 10 mL of
DMF at rt was added 0.232 g (3.57 mmol) of sodium azide. After the
mixture was stirred at 100 °C for 2 h, 10 mL of water was added. The
mixture was extracted three times with 10 mL portions of ether. The
combined organic layers were washed with brine, dried over MgSO4,
and concentrated in Vacuo to produce azide 46 as colorless crystals
suitable for use in the next step without further purification (0.565 g,
1
93%): mp 38-39 °C; H NMR (200 MHz, CDCl3) δ 4.38 (s, 2 H),
6.01 (s, 2 H), 6.88 (s, 1 H), 7.06 (s, 1 H); 13C NMR (75 MHz, CDCl3)
δ 54.4, 102.0, 109.8, 112.9, 114.5, 127.9, 147.6, 148.4; IR (CDCl3)
2899, 2096 cm-1; CIMS m/z (relative intensity) 257 (21), 255 (M+,
21), 215 (100), 213 (M+ - N3, 98); HRMS calcd for C8H6BrN3O2
254.9644, found 254.9654.
Preparation of Cycloadduct 49. To a solution of 399 mg (1.56
mmol) of azide 46 in 16 mL of dry mesitylene under argon was added
410 mg (1.56 mmol) of Ph3P. The mixture was stirred at 50 °C for 4
h. To the above solution was added 736 mg (1.49 mmol) of aldehyde
28 in 4 mL of mesitylene by cannula. After being stirred at 50 °C for
16 h, the mixture was refluxed for 2 h. The mixture was cooled to 0
°C, and 20 mL of THF and 1.49 mL (1.49 mmol) of tetrabutylammo-
nium fluoride (1.0 M solution in THF) were added. After the mixture
was stirred at 0 °C for 1 h, 20 mL of water was added. The mixture
was extracted three times with 20 mL portions of ether. The combined
organic layers were washed with brine, dried over K2CO3, and
concentrated in Vacuo. The residue was purified by flash chromatog-
raphy on silica gel, eluting with ethyl acetate/hexanes (1/19), to produce
amine 49 as a clear oil (535 mg, 63% from aldehyde 28): [R]20
)
D
1
-3.9° (c ) 2.58, benzene); H NMR (300 MHz, CDCl3) δ 0.08 (s, 3
H), 0.077 (s, 3 H), 0.89 (s, 9 H), 1.62-1.72 (m, 1 H), 1.86-2.06 (m,
3 H), 2.11 (d, J ) 2.5 Hz, 1 H), 3.05 (dt, J ) 3.9, 10.5 Hz, 1 H),
3.13-3.20 (m, 1 H), 3.69 (ddd, J ) 2.4, 4.3, 9.8 Hz, 1 H), 3.75 (d, J
) 13.9 Hz, 1 H), 3.83 (d, J ) 13.9 Hz, 1 H), 4.19-4.21 (m, 1 H), 4.55
(d, J ) 11.8 Hz, 1 H), 4.61 (d, J ) 11.8 Hz, 1 H), 5.95 (d, J ) 1.2 Hz,
1 H), 5.96 (d, J ) 1.2 Hz, 1 H), 6.97 (s, 1 H), 7.00 (s, 1 H), 7.25-7.38
(m, 5 H); 13C NMR (75 MHz, CDCl3) δ -5.0, -4.5, 18.1, 25.8 (3 C),
29.0, 29.6, 31.1, 50.5, 51.1, 68.0, 70.5, 71.6, 76.2, 83.8, 101.5, 110.0,
112.6, 113.9, 127.2, 127.4 (2 C), 127.8, 128.2 (2 C), 138.9, 147.2,
147.4; IR (film) 3600-3100, 3080-2800, 2090 cm-1; CIMS m/z
(relative intensity) 574 (13), 572 (M+ + H, 12); HRMS calcd for C29H38-
BrNO4Si 571.1754, found 571.1747.
Preparation of Alkene 50. To a solution of 311 mg (0.545 mmol)
of alkyne 49 in 25 mL of MeOH were added 31.1 mg of Lindlar catalyst
and 5 drops of quinoline. After being stirred under 1 atm of hydrogen
at rt for 19 h, the mixture was filtered through a small plug of Celite
and concentrated in Vacuo. The residue was purified by flash
chromatography on silica gel, eluting with ethyl acetate/hexanes (1/9),
to produce alkene 50 as a white foam (290 mg, 93%): [R]20D ) +1.2°
(c ) 0.48, benzene); 1H NMR (300 MHz, CDCl3) δ 0.03 (s, 3 H), 0.94
(s, 9H), 1.75-1.83 (m, 2 H), 1.90-1.98 (m, 2 H), 2.75-2.81 (m, 1
H), 2.99-3.05 (m, 1 H), 3.65-3.71 (m, 1 H), 3.71 (d, J ) 13.7 Hz, 1
H), 3.78 (d, J ) 13.7 Hz, 1 H), 4.20-4.25 (m, 1 H), 4.58-4.77 (m, 2
H), 5.06-5.18 (m, 2 H), 5.82-5.97 (m, 1 H), 5.96 (s, 2 H), 6.88 (s, 1
H), 7.00 (s, 1 H), 7.27-7.41 (m, 5 H); 13C NMR (75 MHz, CDCl3) δ
-4.8, -4.6, 18.1, 25.9 (3 C), 28.9 (br), 33.4 (br), 38.7 (br), 51.2, 53.7
(br), 68.9, 71.4 (br), 76.4 (br), 101.6, 110.1, 112.7, 114.0, 116.1 (br),
127.1, 127.3 (2 C), 128.1 (2 C), 132.7, 133.0, 139.4, 147.2 (2 C); IR
(film) 3340-3280, 3060-2800 cm-1; EIMS m/z (relative intensity)
573 (M+, 2); HRMS calcd for C29H40BrNO4Si 573.1910, found
573.1878.
Heck Cyclization of Bromo Alkene 50. To a solution of 293 mg
(0.511 mmol) of alkene 50 in 8 mL of dry CH3CN in a sealed tube
were added 177 mg (0.153 mmmol) of Pd(PPh3)4, 0.85 mL (6.13 mmol)
of triethylamine, and 190 mg (1.02 mmol) of trimethylbenzylammonium
chloride. After the mixture was deoxygenated under vacuum three
times, the tube was sealed under argon and heated at 120 °C for 48 h.
After being cooled to rt, the mixture was filtered through a small plug
of Celite and concentrated in Vacuo. The residue was purified by flash
Preparation of Cycloadduct 43. Following the procedure for
preparation of amine 39, aldehyde 31 (27 mg, 0.0547 mmol) was
converted to cycloadduct 43 (14 mg, 52% from 31): 1H NMR (300
MHz, CDCl3) δ 0.05 (s, 6 H), 0.89 (s, 9 H), 1.48-1.54 (m, 2 H), 2.01-
2.13 (m, 1 H), 2.13 (d, J ) 2.4 Hz, 1 H), 2.25-2.35 (m, 1 H), 2.63-
2.71 (m, 1 H), 2.72-2.85 (m, 1 H), 3.41-3.51 (m, 1 H), 3.70-3.87
(m, 3H), 4.64 (d, J ) 12.6 Hz, 1 H), 4.70 (d, J ) 12.6 Hz, 1 H), 5.94
(s, 2 H), 6.74 (d, J ) 7.8 Hz, 2 H), 6.82 (dd, J ) 1.2, 7.8 Hz, 1 H),
6.97 (s, 1 H), 7.23-7.42 (m, 5 H); IR (film) 3300, 2100 cm-1; EIMS
m/z (relative intensity) 493 (M+, 2), 436 (7); HRMS calcd for C29H39-
NO4Si 493.2648, found 493.2625.
Preparation of Azide 46. To a solution of 274 mg (1.19 mmol) of
bromo alcohol 4429b in 5 mL of MeCN at rt under argon was added
357 mg (2.38 mmol) of NaI, followed by slow addition of 0.30 mL
(2.38 mmol) of TMSCl. After being stirred at rt for 15 min, the mixture
was taken up in CH2Cl2 and washed with water, 10% aqueous Na2S2O3

Synthesis of Amaryllidaceae Alkaloids
J. Am. Chem. Soc., Vol. 119, No. 25, 1997 5783
chromatograghy on silica gel, eluting with triethylamine/ethyl acetate/
at rt for 24 h, the mixture was filtered through a small plug of Celite
and concentrated in Vacuo. The crude diol 62 was suitable for use in
the next step without further purification.
hexanes (1/10/90), to produce tetracyclic alkene 51 as a white foam
1
(187 mg, 74%): [R]20 ) -5.5° (c ) 0.73, benzene); H NMR (200
D
MHz, CDCl3) δ -0.02 (s, 6 H), 0.86 (s, 9 H), 1.39-1.62 (m, 3 H),
2.24 (dt, J ) 3.6, 12.5 Hz, 1 H), 2.89-2.99 (m, 1 H), 3.26-3.29 (m,
1 H), 3.56-3.59 (m, 1 H), 3.70 (s, 2 H), 3.94 (ddd, J ) 3.2, 3.6, 11.2
Hz, 1 H), 4.63 (d, J ) 12.4 Hz, 1 H), 4.78 (d, J ) 12.4 Hz, 1 H), 4.88
(d, J ) 2.0 Hz, 1 H), 5.17 (d, J ) 2.0 Hz, 1 H), 5.91 (s, 2 H), 6.60 (s,
1 H), 6.66 (s, 1 H), 7.23-7.41 (m, 5 H); 13C NMR (75 MHz, CDCl3)
δ -4.7, -4.6, 18.2, 25.9 (3 C), 29.9, 38.0, 42.0, 54.2, 62.2, 69.1, 72.2,
77.4, 100.9, 108.3, 109.8, 114.8, 127.1, 127.4 (2 C), 128.1 (2 C), 132.1,
135.5, 139.7, 146.1, 146.3, 153.5; IR (CDCl3) 3066, 2954-2857, 1630,
1612 cm-1; EIMS m/z (relative intensity) 493 (M+, 9); HRMS calcd
for C29H39NO4Si 493.2648, found 493.2629.
To a solution of 1.05 g (8.19 mmol) of naphthalene in 5 mL of dry
DME was added 0.20 g (8.70 mmol) of sodium. The mixture was
stirred at rt for 2 h, during which time the solution turned a deep blue
color. To a solution of the above crude diol 62 in 4 mL of DME at
-78 °C under argon was added the above blue solution dropwise by
cannula until the solution maintained its blue color. The mixture was
quenched with 5 mL of saturated aqueous NaHCO3 solution, and was
extracted three times with 10 mL portions of ethyl acetate. The
combined organic layers were dried over K2CO3 and concentrated in
Vacuo to produce the crude amine 63.
To the above crude amine 63 dissolved in 1.5 mL of CH3CN and 3
mL of ether were added 71 mg (0.271 mmol) of PPh3, 28 mg (0.411
mmol) of imidazole, and 69 mg (0.271 mmol) of I2 at 0 °C under argon.
After being stirred at 0 °C for 15 min, the mixture was quenched with
5 mL of saturated aqueous NaHCO3 solution and 2 mL of saturated
aqueous Na2S2O3 solution, and was extracted three times with 10 mL
portions of ether. The combined organic layers were dried over K2CO3
and concentrated in Vacuo. The residue was purified by flash
chromatography on silica gel, eluting with CH2Cl2/MeOH/NH4OH (97/
2/1), to produce pentacyclic amine 64 as a clear oil (44.8 mg, 82%
from alcohol 54): [R]20D ) +55.4° (c ) 0.16, MeOH); 1H NMR (360
MHz, CDCl3) δ 0.10 (s, 3 H), 0.17 (s, 3 H), 0.89 (s, 9 H), 1.42 (ddd,
J ) 3.9, 12.7, 14.4 Hz, 1 H), 1.57 (ddd, J ) 3.7, 9.3, 14.0 Hz, 1 H),
1.95 (ddd, J ) 2.4, 5.9, 14.4 Hz, 1 H), 2.02 (dd, J ) 6.5, 19.7 Hz, 1
H), 2.53 (d, J ) 2.4 Hz, 1 H), 2.54-2.62 (m, 1 H), 2.90 (d, J ) 11.5
Hz, 1H), 2.92 (s, 1 H), 3.05 (dd, J ) 2.4, 11.4 Hz, 1 H), 3.24 (q, J )
6.7 Hz, 1 H), 3.72 (d, J ) 16.7 Hz, 1 H), 3.86-3.94 (m, 1 H), 4.03-
4.09 (m, 1 H), 4.27 (d, J ) 16.7 Hz, 1 H), 5.878 (d, J ) 1.5 Hz, 1 H),
5.883 (d, J ) 1.5 Hz, 1 H), 6.45 (s, 1 H), 6.50 (s, 1 H); 13C NMR (90
MHz, CDCl3) δ -2.8, -2.2, 18.1, 25.8 (3 C), 32.3, 33.2, 43.6, 45.5,
52.8, 61.0, 61.1, 66.0, 68.0, 100.2, 106.4, 107.1 (2 C), 136.4, 145.7,
146.4; IR (CDCl3) 3500, 3155, 2858-2955, 1794 cm-1; EIMS m/z
Preparation of Toluenesulfonamide 52. To a solution of 125 mg
(0.254 mmol) of amine 51 in 5 mL of pyridine were added 12.5 mg of
DMAP and 97.0 mg (0.508 mmol) of toluenesulfonyl chloride. The
mixture was stirred at 100 °C for 20 h. After the mixture was cooled
to 0 °C, 10 mL of saturated aqueous NaHCO3 solution was added, and
stirring was continued for 15 min. The aqueous layer was extracted
three times with 10 mL portions of ether. The combined organic layers
were washed with brine, dried over MgSO4, and concentrated in Vacuo.
The residue was purified by flash chromatography on silica gel, eluting
with ethyl acetate/hexanes (1/4), to produce toluenesulfonamide 52 as
a white foam (151 mg, 92%): [R]20 ) -13.1° (c ) 0.73, benzene);
D
1H NMR (300 MHz, CDCl3) δ 0.08 (s, 6 H), 0.96 (s, 9 H), 1.42-1.84
(m, 3 H), 2.01-2.11 (m, 1 H), 2.39 (s, 3 H), 2.49-2.55 (m, 1 H),
3.46-3.54 (m, 1 H), 4.11-4.17 (m, 1 H), 4.26-4.34 (m, 2 H), 4.35-
4.50 (m, 1 H), 4.47 (d, J ) 12.3 Hz, 1 H), 4.58 (d, J ) 12.4 Hz, 1 H),
4.98 (s, 1 H), 5.07 (s, 1 H), 5.95 (d, J ) 1.2 Hz, 1 H), 5.96 (d, J ) 1.2
Hz, 1 H), 6.45 (s, 1 H), 6.75 (s, 1 H), 7.18 (d, J ) 7.9 Hz, 2 H),
7.26-7.41 (m, 5 H), 7.53 (d, J ) 7.9 Hz, 2 H); 13C NMR (75 MHz,
CDCl3) δ -4.9, -4.6, 18.2, 21.4, 22.6, 25.8 (3 C), 31.5, 40.7, 47.4,
52.7, 68.3, 70.2, 74.0, 101.1, 108.9 (2 C), 117.0, 127.1 (2 C), 127.47
(2 C), 127.55 (2 C), 128.3 (2 C), 129.4 (2 C), 137.3, 137.7, 138.7,
142.8, 146.8, 147.3, 150.2; IR (CDCl3) 3050-2800, 1599 cm-1; EIMS
m/z (relative intensity) 647 (M+, 0.4), 590 (63); HRMS calcd for C36H45-
NO6SSi 647.2737, found 647.2698.
(relative intensity) 403 (M+, 28), 387 (M+ - OH, 10), 346 (M+
-
t-Bu, 100); HRMS calcd for C22H33NO4Si 403.2179, found 403.2146.
Preparation of Ketone 65. To a solution of 31.0 mg (0.0769 mmol)
of alcohol 64 in 4 mL of dry CH2Cl2 under argon were added 0.5 g of
4 Å molecular sieves and 27.0 mg (0.230 mmol) of N-methylmorpholine
N-oxide at rt. After 10 min, 2.70 mg (0.00769 mmol) of tetrapropyl-
ammonium perruthenate was added. After being stirred at rt for 24 h,
the mixture was filtered through a small plug of Celite and concentrated
in Vacuo. The residue was purified by flash chromatography on silica
gel, eluting with CH2Cl2/MeOH/NH4OH (97/2/1), to produce ketone
65 as a clear oil (29.6 mg, 96%): [R]20D ) -81.7° (c ) 0.28, MeOH);
1H NMR (360 MHz, CDCl3) δ 0.06 (s, 3 H), 0.10 (s, 3 H), 0.85 (s, 9
H), 1.88 (ddd, J ) 3.0, 11.5, 14.4 Hz, 1 H), 2.30-2.36 (m, 1 H), 2.36
(dd, J ) 13.0, 15.6 Hz, 1 H), 2.50 (dd, J ) 5.8, 15.6 Hz, 1 H), 2.62 (d,
J ) 2.6 Hz, 1 H), 2.61-2.70 (m, 1 H), 3.01 (d, J ) 11.9 Hz, 1H), 3.16
(dd, J ) 2.7, 11.9 Hz, 1 H), 3.42-3.53 (m, 1 H), 3.78 (d, J ) 17.6 Hz,
1 H), 3.95 (dd, J ) 3.2, 3.3 Hz, 1 H), 4.32 (d, J ) 17.6 Hz, 1 H), 5.89
(s, 2 H), 6.47 (s, 1 H), 6.49 (s, 1 H); 13C NMR (90 MHz, CDCl3) δ
-2.8, -2.0, 18.2, 25.7 (3 C), 35.4, 40.5, 46.1, 46.9, 51.9, 60.5, 60.6,
71.2, 100.7, 106.5, 107.0, 125.3, 135.4, 145.8, 146.5, 209.1; IR (CDCl3)
3155, 2857-2955, 1794, 1725 cm-1; EIMS m/z (relative intensity) 401
(M+, 6), 344 (M+ - t-Bu, 100); HRMS calcd for C22H31NO4Si
401.2022, found 401.2040.
Preparation of Enone 67. To a solution of 0.19 mL (1.36 mmol)
of diisopropylamine in 4 mL of dry THF at 0 °C under argon was
added 0.54 mL (1.35 mmol) of n-BuLi (2.5 M solution in hexanes).
After 10 min, the solution was cooled to -78 °C. Freshly distilled
TMSCl (0.85 mL, 6.70 mmol) was added to the above solution at -78
°C, followed by dropwise addition of a solution of 27.0 mg (0.0673
mmol) of ketone 65 in 1 mL of THF. After the mixture was stirred
for 5 min, 1.9 mL (13.6 mmol) of triethylamine was added. The
mixture was allowed to warm to 0 °C over 1 h. The mixture was
quenched with 5 mL of saturated aqueous NaHCO3 solution and
extracted three times with 10 mL portions of ether. The combined
organic layers were dried over Na2SO4 and concentrated in Vacuo. The
crude silyl enol ether 66 was immediately used for the next step without
further purification.
Preparation of Alcohol 54. To a solution of 59.0 mg (0.0912 mmol)
of alkene 52 in 4 mL of acetone at -20 °C under argon was added 3.0
mL (0.30 mmol) of dimethyldioxirane33 (0.1 M solution in acetone).
The mixture was stirred at -20 °C for 4 h and was concentrated in
Vacuo. The crude epoxides 55 were produced as a white foam (54
mg, 89%, 2/1 mixture of diastereoisomers determined by the integration
of 1H NMR spectrum of the crude product) and were used for the next
step immediately.
To a solution of 80.0 mg (0.121 mmol) of the above mixture of
epoxides in 5 mL of dry CH2Cl2 at -78 °C under argon was added
58.9 mg (0.363 mmol) of anhydrous FeCl3. After the mixture was
stirred at -78 °C for 30 min, 0.60 mL (0.60 mmol) of DIBALH (1.0
M solution in hexanes) was added at -78 °C. After the mixture was
stirred at 0 °C for 10 min, 5 mL of 5% aqueous HCl solution was
added. The mixture was extracted three times with 10 mL portions of
CH2Cl2. The combined organic layers were washed with brine, dried
over MgSO4, and concentrated in Vacuo. The residue was purified by
flash chromatography on silica gel, eluting with ethyl acetate/hexanes
(1/2), to produce alcohol 54 as a white foam (70.6 mg, 88%): [R]20
D
) +0.9° (c ) 1.05, benzene); 1H NMR (300 MHz, CDCl3) δ 0.01 (s,
3 H), 0.03 (s, 3H), 0.89 (s, 9 H), 1.20-1.36 (m, 1 H), 1.48-1.79 (m,
2 H), 2.32-2.45 (m, 1 H), 2.44 (s, 3 H), 2.49-2.53 (m, 1 H), 3.04-
3.08 (m, 1 H), 3.17-3.21 (m, 1 H), 3.71-3.75 (m, 1 H), 3.83-3.93
(m, 2 H), 4.14-4.17 (m, 1 H), 4.32-4.65 (m, 4 H), 5.94 (s, 2 H), 6.28
(s, 1 H), 6.67 (s, 1 H), 7.27-7.40 (m, 7 H), 7.73 (d, J ) 7.8 Hz, 2 H);
13C NMR (75 MHz, CDCl3) δ -4.9, -4.6, 18.1, 21.5, 25.4, 25.9 (3
C), 29.7, 36.2, 36.8, 43.2, 51.6, 62.4, 68.1, 70.8, 74.7, 101.1, 106.0,
109.4, 127.0 (2 C), 127.5 (2 C), 127.6 (2 C), 128.4 (2 C), 129.8, 132.9,
137.0, 138.6, 139.0, 143.2, 146.1, 147.9; IR (CDCl3) 3622 cm-1; EIMS
m/z (relative intensity) 665 (M+, 1); HRMS calcd for C32H38NO7SSi
(M+ - t-Bu) 608.2138, found 608.2160.
Preparation of Pentacyclic Amine 64. To a solution of 90.0 mg
(0.135 mmol) of alcohol 54 in 15 mL of MeOH was added 18 mg of
10% palladium on carbon. After being stirred under 1 atm of hydrogen

5784 J. Am. Chem. Soc., Vol. 119, No. 25, 1997
Jin and Weinreb
To the above crude 66 dissolved in 3.4 mL of dry MeCN was added
151 mg (0.673 mmol) of Pd(OAc)2 at rt under argon. After being stirred
for 50 h at rt, the mixture was quenched with 5 mL of saturated aqueous
NaHCO3 solution and extracted three times with 5 mL portions of ether.
The combined organic layers were dried over K2CO3 and concentrated
in Vacuo. The residue was purified by flash chromatography on silica
gel, eluting with CH2Cl2/MeOH/NH4OH (97/2/1), to produce 4.7 mg
of starting ketone 65 and 18.0 mg of enone 67 (67% isolated yield,
Preparation of (-)-Montanine (1). To a solution of 8.1 mg (0.018
mmol) of ketal silyl ether 69 in 1 mL of THF at 0 °C was added 0.18
mL (0.18 mmol) of tetrabutylammonium fluoride (1.0 M solution in
THF). After being stirred at rt for 20 h, the mixture was concentrated
in Vacuo. The residue was purified by flash chromatography on silica
gel, eluting with CH2Cl2/MeOH/NH4OH (92.5/5/2.5), to produce ketal
alcohol 71 as a clear oil (6.0 mg, 99%): 1H NMR (300 MHz, CDCl3)
δ 1.70 (ddd, J ) 1.7, 11.9, 12.5 Hz, 1 H), 2.37 (dt, J ) 4.4, 12.5 Hz,
1 H), 2.51 (br s, 1 H), 3.05 (d, J ) 11.1 Hz, 1 H), 3.11 (dd, J ) 1.9,
11.1 Hz, 1 H), 3.24-3.28 (m, 1 H), 3.28 (s, 3 H), 3.32 (s, 3 H), 3.64
-3.74 (m, 1 H), 3.85 (d, J ) 16.8 Hz, 1 H), 4.06-4.11 (m, 1 H), 4.32
(d, J ) 16.8 Hz, 1 H), 5.58 (t, J ) 2.1 Hz, 1 H), 5.88 (d, J ) 1.4 Hz,
1 H), 5.90 (d, J ) 1.4 Hz, 1 H), 6.50 (s, 1 H), 6.55 (s, 1 H); EIMS m/z
(relative intensity) 331 (M+, 100); HRMS calcd for C18H21NO5
331.1420, found 331.1410.
81% based on recovered starting ketone) as clear oils. 67: [R]20
)
D
1
+12.8° (c ) 0.13, MeOH); H NMR (300 MHz, CDCl3) δ 0.04 (s, 3
H), 0.13 (s, 3 H), 0.85 (s, 9 H), 1.90 (ddd, J ) 3.0, 11.8, 12.6 Hz, 1
H), 2.41 (ddd, J ) 2.7, 4.7, 12.6 Hz, 1 H), 3.16 (s, 2 H), 3.42 (s, 1 H),
3.86 (d, J ) 17.1 Hz, 1 H), 3.87-3.96 (m, 1 H), 4.06-4.10 (m, 1 H),
4.40 (d, J ) 17.1 Hz, 1 H), 5.83-5.86 (m, 1 H), 5.93 (d, J ) 1.4 Hz,
1 H), 5.94 (d, J ) 1.4 Hz, 1 H), 6.53 (s, 1 H), 6.59 (s, 1 H); 13C NMR
(75 MHz, CDCl3) δ -5.2, -4.7, 18.2, 25.7 (3 C), 39.7, 46.5, 54.8,
60.0, 60.8, 71.6, 101.0, 106.9, 107.7, 115.7, 124.5, 130.2, 146.3, 147.5,
176.9, 195.5; IR (CDCl3) 3155, 2874-2978, 1794, 1676 cm-1; EIMS
m/z (relative intensity) 399 (M+, 8), 342 (M+ - t-Bu, 100); HRMS
calcd for C22H29NO4Si 399.1866, found 399.1831.
To a solution of 5.0 mg (0.015 mmol) of hydroxy ketal 71 in 1 mL
of dry toluene at 0 °C under argon was added 0.15 mL (0.15 mmol) of
DIBALH (1.0 M solution in hexanes). After the solution was allowed
to warm to rt slowly and stirred for 27 h, 2 drops of saturated aqueous
Na2SO4 solution was added, the mixture was stirred for 30 min, and 2
mL of CH2Cl2/MeOH/NH4OH (92.5/5/2.5) was added. After being
stirred for 30 min, the mixture was filtered, and the filtrate was
concentrated in Vacuo. The residue was purified by flash chromatog-
raphy on silica gel, eluting with CH2Cl2/MeOH/NH4OH (92.5/5/2.5),
to produce starting ketal 71 (0.6 mg, 12%), (-)-coccinine (2) (1.8 mg,
39% isolated yield, 44% based on recovered starting material), and
(-)-montanine (1) (1.9 mg, 41% isolated yield, 47% based on recovered
starting material). (-)-Montanine (1): 1H NMR (500 MHz, CDCl3)
and low- and high-resolution mass spectra were identical to those
Preparation of r-Hydroxy Enone 68. To a solution of 3.0 mg
(0.0075 mmol) of siloxy enone 67 in 1 mL of THF at 0 °C was added
11.3 µL (0.0113 mmol) of tetrabutylammonium fluoride (1.0 M solution
in THF). After being stirred at 0 °C for 2 h and for another 2 h at rt,
the mixture was concentrated in Vacuo. The residue was purified by
flash chromatography on silica gel, eluting with CH2Cl2/MeOH/NH4-
OH (92.5/5/2.5), to produce R-hydroxy enone 68 as a clear oil (2.1
mg, 98%): [R]23546 ) -55.9° (c ) 0.272, MeOH), [R]23D ) -50.7° (c
) 0.272, MeOH) (lit.11 [R]25 ) -47.7° (c ) 0.10, MeOH)). 1H
546
NMR (500 MHz, CDCl3), 13C NMR (125 MHz, CDCl3), and low- and
high-resolution mass spectra were identical to those previously
reported.43
previously reported.12,36 [R]25D ) -83° (c ) 0.06, CHCl3) (lit.3 [R]26
D
) -87.6° (c ) 0.6, CHCl3)); 13C NMR (125 MHz, CDCl3) δ 32.7,
45.6, 55.4, 57.6, 58.7, 60.9, 69.2, 79.8, 100.7, 106.8, 107.3, 112.9, 124.8,
132.5, 145.9, 146.7, 154.3.
Preparation of (-)-Coccinine (2). To a solution of 11.6 mg (0.0291
mmol) of enone 68 in 1 mL of dry MeOH at 0 °C under argon was
added 32 µL (0.289 mmol) of trimethyl orthoformate and 8.3 mg (0.044
mmol) of p-toluenesulfonic acid monohydrate. The solution was slowly
warmed to rt over 2 h and stirred at rt for 2 h. Saturated aqueous
NaHCO3 solution (2 mL) was added, and the mixture was extracted
three times with 5 mL portions of ether. The combined organic layers
were dried over K2CO3 and concentrated in Vacuo. The residue was
purified by flash chromatography on silica gel, eluting with CH2Cl2/
MeOH/NH4OH (97/2/1), to produce ketal 69 as a clear oil (11.8 mg,
91%): 1H NMR (360 MHz, CDCl3) δ 0.07 (s, 3 H), 0.11 (s, 3 H),
0.82 (s, 9 H), 1.78 (dt, J ) 1.8, 12.3 Hz, 1 H), 2.19 (dt, J ) 4.6, 12.3
Hz, 1 H), 3.03 (d, J ) 11.7 Hz, 1 H), 3.08 (dd, J ) 2.0, 11.7 Hz, 1 H),
3.19-3.26 (m, 1 H), 3.25 (s, 6 H), 3.63 -3.72 (m, 1 H), 3.74 (d, J )
16.6 Hz, 1 H), 4.07-4.10 (m, 1 H), 4.31 (d, J ) 16.6 Hz, 1 H), 5.58
(t, J ) 2.0 Hz, 1 H), 5.89 (s, 2 H), 6.46 (s, 1 H), 6.54 (s, 1 H); EIMS
m/z (relative intensity) 445 (M+, 49); HRMS calcd for C24H35NO5Si
445.2284, found 445.2245.
Preparation of Triol 72. To a solution of 21.0 mg (0.0316 mmol)
of alcohol 54 in 4 mL of MeOH was added 25 mg of 10% palladium
on carbon. After being stirred under hydrogen at rt for 20 h, the mixture
was filtered through a small plug of Celite and concentrated in Vacuo.
The residue was dissolved in 3 mL of THF, and 0.063 mL (0.063 mmol)
of tetrabutylammonium fluoride (1.0 M solution in THF) was added.
After being stirred at rt for 3 h, the mixture was concentrated in Vacuo.
The residue was purified by flash chromatography on silica gel, eluting
with CHCl3/MeOH (97/3), to produce triol 72 as a clear oil (14.0 mg,
97%): [R]20 ) -51.9° (c ) 0.91, MeOH); 1H NMR (300 MHz,
D
CDCl3) δ 1.00-1.10 (m, 1 H), 1.34-1.42 (m, 1 H), 1.78 -1.95 (m, 1
H), 2.32-2.42 (m, 1 H), 2.43 (s, 3 H), 3.08-3.14 (m, 1 H), 3.36-
3.40 (m, 1 H), 3.75-3.95 (m, 4 H), 4.11-4.20 (m, 2 H), 4.79 (d, J )
14.6 Hz, 1 H), 5.93 (s, 2 H), 6.46 (s, 1 H), 6.79 (s, 1 H), 7.32 (d, J )
8.0 Hz, 2 H), 7.69 (d, J ) 8.0 Hz, 2 H); 13C NMR (75 MHz, CDCl3)
δ 21.5, 26.9, 34.0, 34.7, 44.6, 54.1, 61.8, 63.4, 67.5, 68.5, 101.1, 105.9,
110.0, 127.2 (2 C), 128.9, 129.9 (2 C), 133.2, 135.5, 143.7, 145.7,
147.5; IR (CDCl3) 3600-3250 cm-1; EIMS m/z (relative intensity) 461
(M+, 22); HRMS calcd for C23H27NO7S 461.1508, found 461.1531.
To a solution of 5.6 mg (0.0126 mmol) of ketal 69 in 1 mL of dry
toluene at rt under argon was added 0.125 mL (0.125 mmol) of
DIBALH (1.0 M solution in hexanes). After the solution was stirred
at rt for 12 h, 2 drops of saturated aqueous Na2SO4 solution was added,
the mixture was stirred for 30 min, and 2 mL of CH2Cl2/MeOH/NH4-
OH (92.5/5/2.5) was added. After being stirred for 30 min, the mixture
was filtered, and the filtrate was concentrated in Vacuo. The residue
was purified by flash chromatography on silica gel, eluting with CH2Cl2/
MeOH/NH4OH (92.5/5/2.5), to produce (-)-coccinine (2) as a clear
Acknowledgment. We thank the National Institutes of
Health (Grant CA-34303) for generous support of this research
and Dr. Stephen P. Fearnley and Professor Raymond. L. Funk
for helpful discussions.
1
oil (3.1 mg, 81%) whose H NMR (500 MHz, CDCl3), and low- and
high-resolution mass spectra were identical to those previously reported:
Supporting Information Available: Experimental details
for the reactions in eqs 2 and 3 and Schemes 4 and 12 (4 pages).
See any current masthead page for ordering and Internet access
instructions.
12,36
[R]25 ) -161° (c ) 0.101, EtOH) (lit.3 [R]27 ) -188.8° (c )
D
D
1.9, EtOH)); 13C NMR (125 MHz, CDCl3) δ 36.0, 46.0, 55.7, 56.7,
58.1, 61.2, 65.6, 77.7, 100.8, 106.9, 107.8, 111.9, 125.0, 132.1, 145.9,
146.8, 156.1. (-)-Montanine (1) (0.3 mg, 8%) was produced as a minor
product.
JA970839N