Total synthesis of (±)-pumiliotoxin C: An electrochemical approach

Article abstract of DOI:10.1002/ejoc.200400846

The total stereoselective synthesis of the decahydroquinoline alkaloid (±)-pumiliotoxin C (cis-195A, 1) is described. The compound was prepared in 15 steps from the commercially available 4-piperidone ethylene ketal 2, in an overall 5% yield. New C-C bond

Full text of DOI:10.1002/ejoc.200400846

FULL PAPER
Total Synthesis of ( )-Pumiliotoxin C: An Electrochemical Approach
Nicolas Girard,[a] Jean-Pierre Hurvois,*[a] Claude Moinet,[a] and Loic Toupet[b]
Keywords: Electrochemistry / Metalation / Natural products / Pumiliotoxin C / Total synthesis
The total stereoselective synthesis of the decahydroquinoline
alkaloid ( )-pumiliotoxin C (cis-195A, 1) is described. The
compound was prepared in 15 steps from the commercially
available 4-piperidone ethylene ketal 2, in an overall 5%
yield. New C–C bonds in the α position relative to the nitro-
gen atom were formed by the metalation of aminonitriles 4
and 15, which were themselves prepared electrochemically.
The ease of this synthesis shows that the N-aryl group is an
efficient nitrogen-protecting group that can be removed in
the last step through a Birch dearomatization. In addition, the
aryl substituent serves as an efficient activator of the nitrogen
atom during the elaboration of aminonitriles 4 and 15. The
oxygen atom of the keto carbonyl group in octahydroquino-
linone 10 was removed by an unprecedented two-step
method involving a Shapiro reaction and a palladium-cata-
lyzed reduction of the intermediate alkene 13. Finally, the
expected stereospecific alkylation–hydride reduction process
of the cis-fused aminonitrile 15 established the trans relation-
ship between the propyl group at C-2 and the methyl sub-
stituent at C-5.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
Introduction
Frogs of the family Dendrobatidae, more commonly
known as “poison-arrow frogs”, produce a vast array of
lipophilic alkaloids, many of which can be isolated from
the glandular secretions of their skin. At least 20 structural
classes have been detected, including histrionicotoxins, epi-
batidine, pyrrolizidines, decahydroquinolines, indolizidines,
and other bicyclic alkaloids. To date, approximately 30 al-
kaloids have been assigned to the decahydroquinoline class;
selected examples are given in Figure 1.[1]
The cis-decahydroquinoline ring is also found in the al-
kaloids of the gephyrotoxin family.[2] Among these alka-
loids, the relatively nontoxic alkaloid pumiliotoxin C (1) is
the most abundant in natural sources, and was the first to
be isolated from the skin secretions of the brightly colored
Panamanian frog Dendrobates pumilio.[3] The difficulty in
isolating more than milligram quantities from natural
sources led chemists to elaborate new chemical pathways
for the formal or total synthesis of 1. For this, one has to
deal with the construction of the cis-decahydroquinoline
ring and should also correctly orient the substituents borne
by the C-2 and C-5 carbon atoms. These synthetic routes
have included alkylation of N-formamidine derivatives,[4]
aminocyclization reactions,[5] Haller–Bauer cleavage,[6] ni-
Figure 1. Structures of cis- and trans-decahydroquinoline alkaloids
from dendrobatid frogs.
trogen fixation,[7] cycloaddition reactions,[8–11] Beckmann
rearrangement of tetrahydroindenones,[12] [3,3]-sigmatropic
rearrangements,[13] addition of Grignard reagents to pyri-
dinium salts,[14] and biomimetic approaches.[15] In addition,
several elegant, well-designed enantioselective syntheses of
natural (–)-1 have been reported in the last decade.[16–20]
In a recent paper,[21] we showed that cis-2,6-dialkylpiperi-
dines III, such as ( )-isosolenopsin A (R1 = C11H23, R2 =
CH3) and ( )-dihydropinidine (R1 = C3H7, R2 = CH3), can
be efficiently synthesized by anodic cyanation of N-phenyl-
2-alkylpiperidine I (Scheme 1).
[a] Laboratoire d’Electrochimie, UMR 6509, Institut de Chimie,
Université de Rennes I,
35042, Campus de Beaulieu, Rennes Cedex, France
Fax: +33-223-235-967
E-mail: Jean-Pierre.Hurvois@univ-rennes1.fr
[b] Groupe Matière Condensée et Matériaux, Université de Rennes
I,
35042, Campus de Beaulieu, Rennes Cedex, France
Fax: +33-223-236-717
E-mail: Loic.Toupet@univ-rennes1.fr
Eur. J. Org. Chem. 2005, 2269–2280
DOI: 10.1002/ejoc.200400846
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2269

N. Girard, J.-P. Hurvois, C. Moinet, L. Toupet
FULL PAPER
Scheme 3. Synthesis of aminonitrile 4. Reagents and conditions:
(a) NaNH2/tBuONa, C6H5Br, THF, 50 °C, overnight, 74%;
(b) – 2 e–, – H+, MeOH, AcOLi (20 gL–1), NaCN (6 equiv.), 70%.
Scheme 1. Synthesis of 2,6-dialkylpiperidines.
The most attractive feature of this sequence is that the
regiochemistry of cyanide introduction in I seems to be well
adapted for the synthesis of III. In other words, a new C–
C bond is created at C-6 regiospecifically, providing a reli-
able route from 2-alkylperidines I to the corresponding 2,6-
dialkyl derivatives III via the bifunctional aminonitrile II.
Our synthetic plan for the stereoselective synthesis of 1 is
outlined in retrosynthesis (Scheme 2). It calls for aminoni-
trile 4 to be elaborated with a masked carbonyl function at
C-4. The unmasked ketone ensures the elaboration of the
bicyclic system, while aminonitrile chemistry affords 2,6-di-
substituted products.
For the synthesis of aminonitrile 4, we used a large-scale
method developed in our laboratory.[23] Before performing
macroscale electrolyses, analytical investigations were car-
ried out with a glassy carbon electrode at a scan rate of
50 mVs–1. Amine 3 (1 gL–1) was dissolved in methanol con-
taining lithium acetate (20 gL–1) as supporting electrolyte
and sodium cyanide (6 equiv. per mol of substrate) as trap-
ping agent. The voltammogram revealed a well-defined
irreversible system at Ep = +0.95 V vs. SCE, followed by
A
an ill-defined peak at Ep = +1.5 V vs. SCE. These data are
in accordance with previBous results obtained in our labora-
tory, which clearly showed that the first oxidation peak is
attributable to the amine–iminium transformation, and the
second oxidation peak corresponds to the oxidation of the
α-aminonitrile system generated at peak A.[23] Taken to-
gether, these results indicate that a selective bielectronic
process could be performed at peak A. Thus, a methanolic
solution of 3 was percolated through a flow cell fitted with
a porous graphite electrode as anode, and the current was
calculated according to a bielectronic process (2 Fmol–1).
After a single passage through the porous anode, the vol-
tammogram recorded for the outlet solution showed the
quasi disappearance of the first oxidation peak. Work-up
and chromatographic purification over a silica column af-
forded the expected aminonitrile 4, which was obtained as
a viscous oil that solidified upon cooling (m.p. 118 °C,
Scheme 2. Retrosynthetic analysis of ( )-pumiliotoxin C (1).
The stereochemical approach was to use the decahydro-
quinoline ring as a rigid template to control the trans rela-
tionship that should be established between the C-5 and C-
2 chiral centers. The role of the cyclic ketal in the formation
of piperidine 4 and in the stabilization of the lithiated ami-
nonitrile (through coordination of Li with O) remained to
be established. In this study, all compounds are racemic,
and consequently, all the stereogenic units will be repre-
sented in a relative configuration.
1
70%). The H NMR spectrum is consistent with a ring cy-
anation and includes a characteristic 2-H signal, which ap-
pears as a broad singlet at δ = 4.67 ppm (1 H).
We set bifunctional aminonitriles 5a,b as initial synthetic
targets of this study (Scheme 4).[24] There are two distinct
processes for the synthesis of 5: deprotonation and alky-
lation of the resulting carbanion. Table 1 details the effects
Results and Discussion
Preparation and Alkylation of Aminonitrile 4
Access to the starting material proved to be straightfor-
ward (Scheme 3). The phenyl group was easily incorporated
by the Caubère coupling of commercially available 4-piperi-
done ethylene ketal 2 with phenyl bromide (1.1 equiv.) in
THF (50 °C, overnight), in the presence of a tBuONa/
NaNH2 mixture as the complex base.[22] Under these condi-
tions, N-phenylpiperidone (3) was obtained in 74% yield
after chromatographic purification.
Scheme 4. Synthesis of aminonitriles 5a,b. Reagents and condi-
tions: (a) LDA (1.1 equiv.), RBr, THF, –20 °C, overnight, 5a (R =
n-C3H7, 75%), 5b [R = (CH2)3Cl, 94%].
2270
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2269–2280

Total Synthesis of ( )-Pumiliotoxin C
FULL PAPER
Table 1. Metalation and alkylation of α-aminonitrile 4.
Entry
Deprotonation, T [°C] / t
Additive
Electrophile
Alkylation, T [°C] / t [h]
Product yield[a][b]
%
,
[h]
1
2
3
4
5
6
7
8
–40 / 2
–40 / 2
–40 / 1
–40 / 5
–20 / 2
–40 / 5
–60 / 1
–40 / 1
–
–
n-C3H7Br
n-C3H7Br
n-C3H7Br
n-C3H7Br
n-C3H7Br
n-C3H7I
–20Ǟ20 / 2
–20Ǟ20 / 12
–20 / 12
–20 / 12
–20 / 12
–20 / 12
–20Ǟ20 / 5
–20 / 12
5a, 40
5a, 55
5a, 60
5a, 10[e]
5a, 75
5a, 72
5a, 75
5b, 94
HMPA[c]
TMEDA[d]
–
–
–
–
n-C3H7I
Br(CH2)3Cl
[a] All products were isolated and characterized by 1H NMR spectroscopy. [b] Yields were determined after column chromatography
through silica gel. [c] HMPA/THF (5:15). [d] 0.2 equiv. [e] Starting compound was recovered in 80% yield.
of temperature, electrophile structure, and solvent additives. tween an aldehyde group, tethered by a three-carbon unit,
All the experiments were carried out with 1.1 equiv. of LDA and C-3 (Scheme 5).
as base and THF as solvent. The deprotonation step was
done at temperatures between –60 °C and –20 °C.
Treatment of the deep-orange anion solution obtained
in this way with 1-bromopropane at –20 °C (no alkylation
occurred below –30 °C), followed by continuous stirring at
room temperature for 2 h (entry 1), gave the desired bifunc-
tional aminonitrile 5a in 40% yield, along with 30–40% of
unreacted aminonitrile 4. Prolonged treatment at room
temperature (12 h, entry 2) did not significantly increase the
yield (55%). Because of these anomalous results, and to
eliminate or minimize the presence of 4, we added HMPA
or TMEDA, solvent additives that are commonly used to
enhance the ionic character of the C–Li bond. Upon com-
paring entries 2 and 3, one sees that the presence of HMPA
as co-solvent had little effect, while the addition of
0.2 equivalents of TMEDA (entry 4) led to an unexpectedly
low 10% yield of 5a accompanied by unreacted aminoni-
trile 4 (80%). Entry 5, however, clearly shows the effect of
the temperature at which the electrophile was condensed
onto the lithiated carbanion. Lowering the temperature to
–20 °C for 12 h significantly improved the formation of
aminonitrile 5a (75%). These differences in reactivity can
be interpreted in terms of electrophile reactivity. At elevated
temperatures (entries 1 and 2), when the lithiated aminoni-
trile reacts with the alkyl halide (RBr) an elimination pro-
cess (E2) occurs competitively. The nature of the electroph-
ile also has an effect as the results (entries 6 and 7) indicate
that the best yields were observed in the presence of iodo-
propane. These results confirm that when the piperidine
ring is oxygenated, the SN/E2 ratio is decreased substan-
tially. Following these data, the alkylation of 4 with 1-
bromo-4-chlorobutane was carried out at –20 °C for 12 h;
the bifunctional aminonitrile 5b was obtained in 94% yield
after careful purification by column chromatography on
silica gel. Having installed the requisite side chain in 5b, we
turned our attention to constructing octahydroquinolinone
10.
Scheme 5. Synthesis of octahydroquinolinone 10. Reagents and
conditions: (a) NaCN (3 equiv.), DMSO, nBu4NI (10 mol-%),
room temp., 48 h, 84%; (b) NaBH4, (4 equiv.), EtOH, room temp.,
overnight, 91%; (c) Dibal-H (1.0 equiv.), CH2Cl2, –60 °C to –20 °C,
overnight, 77%; (d) 1.5 m H2SO4/THF, 78 °C, 5 h, 75%; (e) Me2-
CuLi (2.2 equiv.), Et2O,–15 °C, 30 min, 91%.
This conversion was accomplished by the displacement
of the terminal chlorine atom in 5b by cyanide in DMSO
to afford the polar dinitrile 6 in 84% yield. On further treat-
ment with four equivalents of NaBH4, 6 gave the desired
protected nitrile 7 in a satisfactory 37% yield from 2. The
synthesis of aldehyde 8 was carried out by the mild re-
Synthesis of Octahydroquinolinone 10
For the synthesis of 1, we believed the best synthetic ap- duction of the terminal nitrile group with one equivalent of
proach to be an intramolecular aldol annulation process be- Dibal-H in CH2Cl2 at –20 °C, followed by the hydrolysis of
Eur. J. Org. Chem. 2005, 2269–2280
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2271

N. Girard, J.-P. Hurvois, C. Moinet, L. Toupet
FULL PAPER
the intermediary iminium ion. Unfortunately, the reaction Me derivatives, that conformer A was energetically favored.
proved to be more difficult than anticipated, as it can not Conversely, Booth et al.[27] have observed that when the ni-
be reproduced easily. Thus, the yield of the aldehyde 8 var- trogen atom is substituted by larger groups, conformer B is
ies with the scale of the reaction − it decreased from 95 to preferred. These findings were confirmed by Meyers et
55% starting from an increasing amount of nitrile 7 (0.70 al.,[4] who showed that in the case of N-formamidines and
1
to 7.06 mmol). The H NMR spectrum of 8 is consistent N-Boc derivatives, the cis-decahydroquinoline system exists
with the structure assigned. Specifically, the aldehydic pro- predominantly in conformation B. These results agree with
ton displays a characteristic triplet (J = 1.6 Hz) signal at δ = our observations, and support the fact that 10 would exist
9.69 ppm (1 H). The spectrum also contains a characteristic in conformation B with selectivity greater than 99%, as de-
1
multiplet system at δ = 3.93–4.04 ppm (4 H), attributable to termined by H NMR spectroscopy. Taken together, these
the ethylene ketal moiety.
results also indicate that the conjugate addition of Me2CuLi
For a simple and efficient synthesis of enone 9, it was to enone 9 proceeds in an efficient stereocontrolled mode
clear that both the ketone unmasking and the annulation (path a, Scheme 6).[28] The addition reaction between Me2-
process should be done in the same reaction vessel. To this CuLi (Z) and the ethylenic double bond in conformer 9A
end, the exposure of 8 to acidic conditions was investigated. yields the trans-axial/equatorial enolate 11A, while the sim-
Heating 8 in refluxing THF for 3 h in the presence of 1.5 m ilar addition to conformer 9B (path aЈ) leads to the energet-
HCl resulted in the formation of 9 in a modest 30% yield, ically disfavored enolate 11B, which suffers from unfavor-
accompanied by several polar side-products that were easily
separable by column chromatography. By using oxygen-free
solvents and heating 8 for 5 h in a mixture containing
H2SO4 in place of HCl, enone 9 was obtained in an im-
proved 75% yield. Structural determination of 9 was
straightforward. In the 13C NMR spectrum, the C-4 and C-
5 carbon atoms give resonance signals at δ = 199.09 and
1
138.18 ppm. In the H NMR spectrum recorded in CDCl3,
5-H exhibits a characteristic narrow multiplet signal at δ =
6.84 ppm (1 H).
For the stereoselective introduction of the C-5 methyl
group in 9, we turned to the conjugate addition of an or-
ganocuprate reagent, as reported by Comins[14] and
Kunz.[18] The treatment of 9 with an ethereal solution con-
taining 2.2 equiv. of Me2CuLi at –15 °C for 30 min led, af-
ter the addition of an excess of water to the resulting eno-
late 11, to the expected adduct 10, with an excellent dia-
stereoselectivity (de Ϸ 99%), in 91% yield (Scheme 6). The
13C NMR spectrum of 10 exhibits a set of 14 resonance
lines, two of which are due to the C-4a and C-5 carbon
atoms at δ = 55.67 and 26.78 ppm, respectively. The 1H
NMR spectrum of 10, well resolved, allowed the complete
assignment of all the signals. The most salient feature is the
doublet signal (J = 7.4 Hz) at δ = 1.06 ppm (3 H) corre-
1
sponding to the 5-Me group. The H-1H COSY spectrum
shows that 8-Ha exhibits a well-defined quadruplet of
Scheme 6. Stereochemical pathway for the addition of Me2CuLi to
doublets (2J = 3J = 11.6, 3J = 3.5 Hz, 1 H) at δ = 1.37 ppm.
hexahydroquinolinone 9. For sake of simplicity only the cyclohexyl
ring is represented.
Since 8-Ha exhibits two large coupling constants with its
neighbors, it was concluded that it is in an axial position.
These data allowed us to determine the stereochemistry of
10. A NOESY experiment showed correlations between 5-
Me, 4a-H, and 8a-H, which are typical of a cis-fused deca-
hydroquinoline class, while a correlation between 2-Ha (δ =
3.50 ppm) and 8-Ha (δ = 1.37 ppm) indicated that these two
protons are close (Figure 2).[25] Because this spatial rela-
tionship is possible only in conformer B, it was assumed
that 10 had a gephyrotoxin-like conformation. At first sight,
this result seemed to be contradictory. Eliel et al.[26] have
extensively studied the conformational equilibrium of vari-
ous substituted cis-decahydroquinolines. On the basis of
Figure 2. Conformational study of octahydroquinolinone 10. Char-
13C NMR studies, they concluded, in the case of N-H or N- acteristic NOE effects are indicated.
2272
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2269–2280

Total Synthesis of ( )-Pumiliotoxin C
FULL PAPER
able 1,3-diaxial interactions between the R group and the of nBuLi in THF followed by stirring for 2 h at room tem-
methyl substituent.[29]
perature, upon which N2 evolved. The addition of water to
the resulting vinyllithium salt yielded alkene 13 in a satisfac-
1
tory 82% yield. Both the H and 13C NMR spectra indi-
cated the presence of a single product, providing evidence
that the double bond had formed regiospecifically. From
the literature data,[32] this result was predictable and may
arise from the syn-dianion effect. Several studies have
shown that in ether solvents the geometry of the tosylhydra-
zone directs the regioselectivity of alkene formation. In the
1H NMR spectrum of 13, recorded in CDCl3, the ethylenic
protons are found at δ = 5.70 (1 H) and 5.88 ppm (1 H) as
an AB resonance system (JA,B = 10.1 Hz). Interestingly, the
Shapiro reaction is a particularly attractive method for elab-
orating a Δ3-piperidine system, which upon functionali-
zation of the alkene double bond enables the preparation
of polysubstituted decahydroquinolines.
Completion of the Synthesis of ( )-Pumiliotoxin C
Synthesis of Aminonitrile 15
For the synthesis of decahydroquinoline 1, we thought to
remove the undesired carbonyl group at C-4 (Scheme 7).
Removal can be achieved classically by the Wolff–Kishner
reduction or, more efficiently, by two-step methods involv-
ing the catalytic reduction of a vinyl triflate[14] or the desul-
furization of a dithiolane[18] in the presence of Raney nickel.
For our purposes, the alkene double bond in 13 was re-
duced by a palladium-catalyzed hydrogenation, affording
decahydroquinoline 14 in 85% yield. Because it appeared
difficult to stereoselectively introduce a propyl chain at the
remaining carbon center (C-2),[4] we reasoned that the syn-
thesis of aminonitrile 15 should constitute an attractive
alternate process (Scheme 7). Our operating strategy rested
on the assumption that 15 should be synthesized regioselec-
tively, thus permitting the further incorporation of the req-
uisite n-propyl chain through a nitrile-stabilized carbanion.
Interestingly, the voltammogram of a methanolic solution
of amine 14 displays a characteristic irreversible system at
Ep = +0.95 V vs. SCE, once more indicating the formation
of an iminium ion at the anode. Consequently, 14 was elec-
trolyzed in an undivided batch cell at controlled potential
(Ep = +0.80 V vs. SCE) in a similar medium to that used
for the electrolysis of 3. After the consumption of two Fara-
days per mol of substrate, aminonitrile 15 was obtained in
a modest 45% yield as a mixture of two isomeric species.
1
The H NMR spectrum of that mixture revealed the pres-
ence of epimers at C-2 in a 6:4 ratio. Characteristic low-
field signals attributed to the 2-H protons of each epimer
are found at δ = 4.17 and 4.40 ppm, whereas the 8a-H pro-
ton signals resonate at δ = 3.55 and 4.01 ppm. Interestingly,
this product distribution indicates that the formation of the
intermediary iminium ion at the anode is highly regioselec-
tive. The presence of single adducts cyanated at C-2 reveals
that favorable interactions occur between the partially occu-
pied nitrogen lone-pair of the intermediary radical cation
and the vicinal methylene protons at C-2. Hence, cyanation
takes place at the less-substituted carbon atom.[33,34] Note
Scheme 7. Synthesis of aminonitrile 15. Reagents and conditions:
(a) TsNHNH2 (1.1 equiv.), EtOH, 78 °C, overnight, 83%; (b)
nBuLi (2.2 equiv.), THF, –60 °C, 90 min, room temp., 2 h, 82%; (c)
Pd/C (10%), MeOH, H2 (1 atm.), 10 h, 85%; (d) – 2 e–, – H+,
MeOH, AcOLi (20 gL–1), NaCN (6 equiv.), 78%.
A new solution to this problem is the Shapiro reac- that no detectable amount of the alternate regioisomer,
tion[30,31] followed by the catalytic reduction of an alkene which would result from a proton loss at C-8a, has ever
derivative. Thus, heating 10 in refluxing ethanol, in the pres- been encountered.
ence of 1.1 equiv. of tosylhydrazine, afforded tosylhydra-
To improve the yield of 14, several experiments were un-
zone 12 (83%) as a white solid which melts at 182–184 °C. dertaken. In one of these, we found that when the electroly-
The synthesis should give one geometric isomer in which sis was carried out at a planar vitreous electrode in a divided
the carbon–nitrogen double bond probably has an (E) con- cell on a 2.80 mmol scale the yield of 15 reached 78%.
figuration; no efforts were made to determine the geometry These encouraging observations prompted us to incorpo-
of this product. The Shapiro reaction was performed by rate the propyl chain at C-2 through the metalation of the
treatment of a THF solution of 12 at –60 °C with 2.2 equiv. aminonitrile function of 15.
Eur. J. Org. Chem. 2005, 2269–2280
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2273

N. Girard, J.-P. Hurvois, C. Moinet, L. Toupet
FULL PAPER
Synthesis of ( )-Pumiliotoxin C
A subsequent X-ray study of these crystals provided the
final proof that our predictions were correct, as in the pre-
dominant conformation the propyl chain attached to C-2
and the methyl group attached to C-5 are equatorial, and
hence mutually trans. In addition, aminonitrile 16 has the
required pumiliotoxin-like conformation. The success of the
alkylation lies also in the fact that both cis- and trans-15
yield the same lithiated species A, which is attacked from
the equatorial direction by the entering electrophile to pro-
duce the energetically favored reactant-like transition state
B (Scheme 9). It should also be pointed out that in transi-
tion state B, a maximum orbital overlap between the axial
cyano group and the nitrogen lone-pair is maintained.[35]
The alkylation of a cis/trans mixture (40:60) of 15 in
THF was performed by adding 1.5 equiv. of LDA, and then
by condensing the resulting anion solution with propyl bro-
mide at –20 °C overnight (Scheme 8). Workup yielded the
bifunctional aminonitrile 16 (72%) as a single adduct (de Ϸ
99%). The spectroscopic data (1H, 13C NMR) of this ad-
duct are consistent with the formation of one geometric iso-
mer, although the relative orientation of the side-chain sub-
1
stituent could not be determined from the H NMR spec-
trum. Interestingly, single crystals were obtained by a slow
crystallization of 16 from a mixture of diethyl ether and
petroleum ether (Figure 3).
Scheme 9. Alkylation of aminonitrile 15.
In the penultimate step, reduction of the aminonitrile
system in 16 should occur with retention of configuration
at C-2. Reasoning that in conformation D a severe synaxial
interaction between the C-8–C-8a bond and the propyl
chain borne by C-2 should place the intermediate iminium
system in conformation C, we planned the hydride re-
duction of 16 (Scheme 10). Thus, treatment of 16 with an
excess of NaBH4 in ethanol resulted in the synthesis of de-
Scheme 8. Synthesis of ( )-pumiliotoxin C (1). Reagents and condi-
tions: (a) LDA (1.5 equiv.), n-C3H7Br, THF, –20 °C, overnight,
72%; (b) EtOH, NaBH4, (4.0 equiv.), room temp., overnight, 81%;
(c) Li (130 equiv.), liq. NH3/THF/EtOH (20:10:8), –40 °C, 1 h,
H2SO4/H2O/EtOH (0.5:5.0:4.5), 60 °C, 10 min, 78%.
Figure 3. PLATON view of aminonitrile 16.
Scheme 10. Hydride reduction of aminonitrile 16.
2274
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2269–2280

Total Synthesis of ( )-Pumiliotoxin C
FULL PAPER
cahydroquinoline 17 in 81% yield. Attack from the bottom
face of the molecule, which would have generated the less
thermodynamically stable isomer of 17, was not observed.
1
Surprisingly, both the H and 13C NMR spectra of 17
display unresolved coalescent signals. For example, in the
1H NMR spectrum, recorded in C6D6, diagnostic protons
2-H and 8a-H resonate as broadened singlets at δ = 3.00 (1
H) and 3.29 ppm (1 H), respectively. This anomalous be-
havior was also found in the broad-band decoupled 13C
NMR spectrum (C6D6, 300 K), in which the C-3Ј, C-2, and
C-8a carbon atoms (Scheme 8) resonate as ill-defined sys-
tems at δ = 124.38, 60.90, and 59.44 ppm, respectively.
However, heating the sample at 343 K induced decoalesc-
ence of these broad peaks and a sharpening of all the sig-
nals. We assessed our product ratio to be at least 99:1, but
we were not able to determine the stereochemistry of 17 by
spectroscopic means. After a few days, amine 17 crystallized
as colorless plates (m.p. Ͻ 40 °C). An X-ray study of these
plates revealed the stereochemical outcome of the reduction
process. The ORTEP view clearly shows that the propyl
chain is equatorial (Figure 4), indicating that the hydride
anion was incorporated as shown in Scheme 10. Finally, the
N-aryl bond in 17 was cleaved by the reductive dearomati-
zation of the phenyl substituent in a single electron transfer.
In a preceding report,[21] we noticed that reduction of the
phenyl group of several N-phenyl-2,6-dialkylpiperidines was
possible in a NH3/THF/EtOH (20:10:4) mixture in the pres-
ence of 10 equiv. of Li. Under these reaction conditions,
amine 17 was not completely reduced and a mixture of un-
stable enamines 18 and 19, which were rapidly hydrolyzed
in sulfuric acid medium (H2SO4/H2O/EtOH, 0.5:4.5:5.0,
Scheme 11), was obtained. Note that when ethanol was
used as proton donor, over-reduction of an ethylenic double
bond occurred (step c, Scheme 11) to yield substantial
amounts of enamine 19 (30–40%).
Scheme 11. Birch reduction of decahydroquinoline 17. (a) + e–; (b)
+ 2 H+, + e–; (c) + 4 H+, + 3 e–; (d) H2SO4/H2O/EtOH (0.5:5.0:4.5),
60 °C, 10 min.
Workup afforded a crude mixture, which was purified by
careful filtration through silica gel with diethyl ether and
diethyl ether saturated with gaseous ammonia as successive
eluents. The less-polar unreacted amine 17 eluted first, fol-
lowed by 1, which was isolated in a modest 25% yield. Nev-
ertheless, the high-resolution mass spectrum supported the
exact formula C13H25N, with M+ at 195.1982. Realizing
that protonation of the intermediate radical anion E (Steps
b and c in Scheme 11) should constitute the driving force
of the reduction process, we attempted to increase the effi-
ciency of the Birch reduction. Thus, increasing the amount
of ethanol twofold and performing the reduction with
130 equiv. of Li allowed us to prepare 1 in a much improved
1
78% yield, whose high-field H NMR spectrum confirmed
the molecular formula, and whose 13C NMR spectrum was
consistent with the literature values.[8]
Conclusions
In summary, an efficient, scalable, stereocontrolled syn-
thesis of ( )-pumiliotoxin C, adaptable to the synthesis of
an interesting set of decahydroquinolines, has been devel-
oped. Apart from the generally high yields in the individual
steps, the most prominent features of the synthesis are (a)
the viability of the aryl group as an efficient activator and
protecting group of the nitrogen atom, (b) the unprece-
dented utilization of a regioselective Shapiro reaction to re-
move the carbonyl group embedded in a decahydroquino-
line ring, and (c) the efficiency of α-aminonitrile chemistry
for the α functionalization of 4-piperidone systems.
Experimental Section
General Techniques: Purification by column chromatography was
Figure 4. ORTEP drawing of decahydroquinoline 17.
performed with 70_230 mesh silica gel (Merck). TLC analyses were
Eur. J. Org. Chem. 2005, 2269–2280
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2275

N. Girard, J.-P. Hurvois, C. Moinet, L. Toupet
FULL PAPER
carried out on alumina sheets precoated with silica gel 60 F254 and
NMR (CDCl3, 50 MHz): δ = 35.35, 37.31, 45.48, 50.75, 64.99,
visualized with UV light; Rf values are given for guidance. IR spec-
65.27, 105.49, 117.77, 119.18, 123.02, 129.88, 149.25 ppm. HRMS
tra were recorded with a Perkin–Elmer FT-IR 16PC (KBr powder (C14H16N2O2 [M+]): calcd. for 244.1212; found 244.1210.
or dichloromethane). NMR spectra were recorded with a Bruker
Avance DRX 500 FT spectrometer [500 MHz (1H) and 100 MHz
(13C)], a Bruker AH 300 FT spectrometer [300 MHz (1H) and
75 MHz (13C)], or a Bruker DPI 200 FT [200 MHz (1H) and
50 MHz (13C)]. Chemical shifts are expressed in ppm downfield
from TMS; coupling constants (J) are given in Hertz. High resolu-
tion mass spectra were obtained with a Mat 311 double focusing
instrument at the CRMPO with a source temperature of 170 °C.
An ion accelerating potential of 3 kV and ionizing electrons of
70 eV were used. For air-sensitive reactions, all glassware was oven
dried (120 °C) over a 24 h period and cooled under a stream of
argon. All commercially available reagents were used as supplied.
THF was distilled from sodium benzophenone ketyl and stored
under an argon atmosphere. Methylene chloride and tert-butanol
were distilled from calcium hydride. Diisopropylamine was distilled
from potassium hydroxide. Yields refer to chromatographically and
spectroscopically (1H, 13C) homogeneous material. Air-sensitive
reagents were transferred by syringe or with a double-ended needle.
C14H16N2O2: calcd. C 68.83, H 6.60, N 11.48; found C 68.81, H
6.52, N 11.23.
8-Phenyl-7-propyl-1,4-dioxa-8-azaspiro[4.5]decane-7-carbonitrile
(5a): nBuLi (1.6 m in hexane; 1.4 mL, 2.24 mmol) was added (by
syringe) to a THF (5.0 mL) solution of diisopropylamine (0.38 mL,
0.27 g, 2.66 mmol) at –80 °C. The solution was stirred at that tem-
perature for 15 min and was then allowed to warm to 0 °C over a
30 min period. The resulting LDA solution was transferred (with a
double-ended needle) to a THF (10 mL) solution of 4 (0.5 g,
2.05 mmol) cooled to –55 °C. The solution was warmed to –20 °C
for 2 h, and 1-bromopropane (0.22 mL, 0.30 g, 2.44 mmol) was
added dropwise. The mixture was maintained at –20 °C for 12 h,
and the reaction mixture was diluted with water (100 mL). The
aqueous phase was extracted with diethyl ether (100 mL×3), the
combined organic layers were dried with MgSO4, and the solvents
evaporated in vacuo. The oily residue was purified by column
chromatography (diethyl ether/petroleum ether, 45:55) to afford 5a
(0.44 g, 75%) as pale-yellow oil. Rf (diethyl ether/petroleum ether,
50:50) = 0.50. 1H NMR (CDCl3, 300 MHz): δ = 0.78 (t, J = 6.8 Hz,
3 H), 1.28–1.50 (m, 4 H), 1.79–1.84 (m, 2 H), 1.93 (td, J = 13.0,
5.0 Hz, 1 H), 2.13 (dd, J = 13.5, 2.5 Hz, 1 H), 3.59 (dm, J = 9.0 Hz,
1 H), 3.57 (td, J = 12.5, 2.8 Hz, 1 H), 3.92–4.11 (m, 4 H), 7.19–
7.37 (m, 5 H) ppm. 13C NMR (CDCl3, 75 MHz): δ = 14.00, 17.05,
35.68, 42.12, 43.25, 50.08, 59.82, 64.32, 64.82, 105.86, 120.44,
126.68, 127.36, 129.00, 148.07 ppm.
8-Phenyl-1,4-dioxa-8-azaspiro[4.5]decane (3): A THF (10 mL) solu-
tion containing tBuOH (2.59 g, 35 mmol) was added dropwise (by
syringe) to
a THF (30 mL) suspension of NaNH2 (4.10 g,
105 mmol). The resulting mixture was stirred at room temperature
for 1 h under an atmosphere of argon. 4-Piperidone ethylene ketal
(2; 5.0 g, 35.0 mmol) and bromobenzene (6.03 g, 38.4 mmol) were
added sequentially over a 15 min period. The solution was stirred
at 50 °C overnight and the crude reaction mixture was quenched
by the addition of an excess of water. The solution was extracted
with diethyl ether (50 mL×3). The combined organic layers were
extracted with a 10% HCl solution (50 mL), which was made basic
by the addition of NaOH pellets. The aqueous layer was extracted
with diethyl ether (50 mL×3) and the ethereal phases were dried
with MgSO4 and concentrated in vacuo to afford an oil, which
was purified by column chromatography on silica gel (diethyl ether/
petroleum ether, 35:65) to afford 3 (5.69 g, 74%) as a pale-yellow
7-(3-Chloropropyl)-8-phenyl-1,4-dioxa-8-azaspiro[4.5]decane-7-car-
bonitrile (5b): LDA (1.5 m in cyclohexane; 16.5 mL, 24.75 mmol)
was added (by syringe) to a THF (35.0 mL) solution of 4 (5.50 g,
22.54 mmol) cooled to –40 °C. The solution was stirred at that tem-
perature for 1 h and 1-bromo-3-chloropropane (2.45 mL, 3.90 g,
24.80 mmol) was added dropwise. The mixture was maintained at
–20 °C over a 12 h period, and was then diluted with water
(100 mL). The aqueous phase was extracted with diethyl ether
(100 mL×3) and the combined organic layers were dried with
MgSO4 and evaporated under reduced pressure. The oily residue
was purified by column chromatography (diethyl ether/petroleum
ether, 50:55) to afford 5b (6.77 g, 94%) as a pale-yellow oil. Rf
oil. R (diethyl ether/petroleum ether, 50:50) = 0.65. IR (KBr): ν =
˜
f
1228, 1497, 1599, 2830, 2884, 2958 cm–1. 1H NMR (CDCl3,
200 MHz): δ = 1.82 (t, J = 5.8 Hz, 4 H), 3.31 (t, J = 6.0 Hz, 4 H),
3.97 (s, 4 H), 6.82 (t, J = 7.2 Hz, 1 H), 6.93 (d, J = 7.9 Hz, 2 H),
7.24 (t, J = 7.3 Hz, 2 H) ppm. 13C NMR (CDCl3, 50 MHz): δ
= 35.35, 48.19, 64.98, 107.63, 117.06, 119.17, 129.56, 151.43 ppm.
HRMS (C13H17NO2 [M+]): calcd. for 219.1259; found 219.1264.
(diethyl ether/petroleum ether, 40:60) = 0.27. IR (KBr): ν = 1105,
˜
1142, 1227, 1365, 1495, 1598, 2831, 2883, 2957 cm–1. 1H NMR
(CDCl3, 200 MHz): δ = 1.57–1.98 (m, 7 H), 2.07 (dd, J = 13.5,
2.3 Hz, 1 H), 3.00 (dm, J = 13.5 Hz, 1 H), 3.25–3.41 (m, 2 H), 3.52
(td, J = 12.3, 3.2 Hz, 1 H), 3.90–4.11 (m, 4 H), 7.17–7.36 (m, 5 H)
ppm. 13C NMR (CDCl3, 50 MHz): δ = 27.37, 36.09, 37.86, 43.74,
44.73, 50.71, 59.75, 64.84, 65.31, 106.16, 119.99, 127.34, 127.70,
129.64, 148.26 ppm. HRMS (C17H21ClN2O2 [M+]): calcd. for
320.1292; found 320.1269. C17H21ClN2O2: calcd. C 63.65, H 6.60,
N 8.73; found C 63.30, H 6.66, N 8.35.
8-Phenyl-1,4-dioxa-8-azaspiro[4.5]decane-7-carbonitrile (4): Com-
pound 3 (6.9 g, 31.5 mmol) was dissolved in 2.0 L of methanol con-
taining lithium acetate (20 gL–1) and sodium cyanide (9.3 g,
190.0 mmol). The solution was filtered through a Millipore system
(5 μm) and electrolyzed in a flow cell fitted with a graphite-felt
anode (diameter 50 mm, thickness 12 mm). The flow rate of the
solution (f) was regulated by a peristaltic pump at 5 mLmin–1 and
the current (iox = 253 mA) calculated[36] according to a bielectronic
process. The outlet solution was concentrated under reduced pres-
sure and the crude material was taken-up in water. The aqueous
phase was extracted with dichloromethane (50 mL×3) and the or-
ganic layers were dried with MgSO4 and concentrated. The crude
material was purified by column chromatography (dichlorometh-
ane/petroleum ether, 85:15) to yield aminonitrile 4 (5.35 g, 70%)
as a white powder, m.p. 118 °C (diethyl ether). Rf (diethyl ether/
7-(3-Cyanopropyl)-8-phenyl-1,4-dioxa-8-azaspiro[4.5]decane-7-car-
bonitrile (6): Compound 5b (1.94 g, 6.06 mmol) and NaCN (0.89 g,
18.16 mmol) were dissolved in DMSO (25 mL) in the presence of
tetrabutylammonium iodide (10 mol-%). The reaction mixture was
stirred at room temperature for 48 h, and was then quenched by
the addition of water (100 mL). The product was extracted with
diethyl ether (50 mL×3), dried with anhydrous MgSO4, and con-
centrated under reduced pressure. The crude oil was purified by
petroleum ether, 50:50) = 0.24. IR (KBr): ν = 1369, 1494, 1600, column chromatography (diethyl ether/petroleum ether, 15:85) to
˜
2226, 2836, 2894, 2973 cm–1. 1H NMR (CDCl3, 200 MHz): δ =
1.85–2.27 (m, 4 H), 3.28–3.53 (m, 2 H), 3.96–4.13 (m, 4 H), 4.66–
4.69 (m, 1 H), 6.98–7.36 (m, 3 H), 7.25–7.35 (m, 2 H) ppm. 13C
afford 6 (1.59 g, 84%) as a highly viscous oil. Rf (petroleum ether/
diethyl ether, 30:70) = 0.26. IR (KBr): ν = 704, 774, 1143, 1492,
˜
1630, 2245, 2960 cm–1. 1H NMR (CDCl3, 300 MHz): δ = 1.56–1.93
2276
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2269–2280

Total Synthesis of ( )-Pumiliotoxin C
FULL PAPER
(m, 7 H), 2.08 (dd, J = 13.5, 2.5 Hz, 1 H), 2.10–2.30 (m, 2 H), 3.00
chromatography (petroleum ether/diethyl ether, 40:60) to yield 9
(dm, J = 12.0 Hz, 1 H), 3.54 (td, J = 12.0, 2.8 Hz, 1 H), 3.92–3.97 (0.45 g, 75%) as a slightly yellow oil. Rf (petroleum ether/diethyl
(m, 2 H), 4.01–4.10 (m, 2 H), 7.20–7.37 (m, 5 H) ppm. 13C NMR
ether, 40:60) = 0.37. IR (KBr): ν = 700, 750, 1600, 1690, 2960,
˜
1
(CDCl3, 75 MHz): δ = 17.00, 20.05, 35.59, 38.76, 43.28, 50.23, 2930, 3055 cm–1. H NMR (CDCl3, 300 MHz): δ = 1.47 (qd, J =
59.22, 64.39, 64.86, 105.58, 118.77, 119.85, 127.06, 127.19, 129.30, 10.60, 4.60 Hz, 1 H), 1.55–1.70 (m, 1 H), 1.86 (dt, J = 10.70,
147.68 ppm. HRMS (C18H21N3O2 [M+]): calcd. for 311.1634;
found 311.1643. C18H21N3O2: calcd. C 69.43, H 6.80, N 13.49;
found C 69.80, H 6.86, N 12.97.
4.0 Hz, 1 H), 1.90–2.01 (m, 1 H), 2.28–2.35 (m, 2 H), 2.62 (t, J =
6.10 Hz, 2 H), 3.39 (dt, J = 13.0, 6.0 Hz, 1 H), 3.63 (dt, J = 13.0,
5.5 Hz, 1 H), 4.04–4.10 (m, 1 H), 6.83–6.85 (m, 1 H), 6.95–7.03 (m,
3 H), 7.28–7.34 (m, 2 H) ppm. 13C NMR (CDCl3, 75 MHz): δ =
20.40, 25.86, 27.82, 39.70, 45.98, 58.05, 118.91, 121.15, 129.28,
137.79, 138.18, 149.45, 199.09 ppm. HRMS (C15H17NO [M+]):
calcd. for 227.1310; found 227.1312. C15H17NO: calcd. C 79.26, H
7.54, N 6.16; found C 79.58, H 7.55, N 5.91.
4-(8-Phenyl-1,4-dioxa-8-azaspiro[4.5]dec-7-yl)butanenitrile
(7):
NaBH4 (0.70 g, 18.50 mmol) was slowly added to an ethanol solu-
tion (20 mL) of 6 (1.43 g, 4.60 mmol). After the addition had been
completed (approximately 15 min), the resulting mixture was
stirred overnight and the solvent was then evaporated to dryness.
The crude material was taken-up with a 15% ammonia solution
(50 mL) and extracted with diethyl ether (50 mL×3). The aqueous
layer was removed and the combined organic phases were dried
with anhydrous MgSO4. The crude mixture was purified by column
chromatography (petroleum ether/diethyl ether, 40:60) to afford 7
(1.20 g, 91%) as a colorless oil. Rf (petroleum ether/diethyl ether,
(4aS*,5R*,8aR*)-5-Methyl-1-phenyloctahydroquinolin-4(1H)-one
(10): MeLi (1.5 m in diethyl ether; 17.8 mL, 26.70 mmol) was added
to a cooled (–30 °C), stirred suspension of CuI (2.55 g, 13.40 mmol)
in diethyl ether (30 mL). This mixture was stirred for 1 h, and then
a solution of enone 9 (1.38 g, 6.08 mmol) in diethyl ether (20 mL)
was added with a cannula and the reaction mixture was allowed to
30:70) = 0.42. IR (KBr): ν = 755, 1141, 1499, 1597, 2244, warm up to –15 °C. After 30 min, TLC monitoring showed the
˜
1
2957 cm–1. H NMR (CDCl3, 300 MHz): δ = 1.51–1.71 (m, 2 H), complete disappearance of starting material. The reaction mixture
1.72–1.85 (m, 5 H), 1.96 (dd, J = 13.5, 5.4 Hz, 1 H), 2.24 (t, J =
7.2 Hz, 2 H), 3.15–3.28 (m, 1 H), 3.42 (dt, J = 13.0, 3.4 Hz, 1 H),
3.83 (q, J = 4.7 Hz, 1 H), 3.91–4.03 (m, 4 H), 6.86 (t, J = 7.3 Hz,
was quenched by the addition of a 10% NH3 solution (50 mL) and
was then extracted with dichloromethane (30 mL×3). The com-
bined organic layers were dried with MgSO4 and concentrated in
1 H), 6.95 (d, J = 7.9 Hz, 2 H), 7.25 (t, J = 7.0 Hz, 2 H) ppm. 13C vacuo. The crude yellow oil was purified by column chromatog-
NMR (CDCl3, 75 MHz): δ = 17.17, 22.84, 29.49, 33.95, 36.94,
43.32, 55.61, 63.91, 64.67, 107.42, 117.79, 119.74, 119.78, 129.30,
150.37 ppm. HRMS (C17H22N2O2 [M+]): calcd. for 286.1681;
found 286.1687. C17H22N2O2: calcd. C 71.30, H 7.74, N 9.78;
found C 71.36, H 7.81, N 9.57.
raphy to afford 10 (1.35 g, 91%) as a slightly amber oil. Rf (petro-
leum ether/diethyl ether, 70:30) = 0.43. IR (KBr): ν = 740, 1265,
˜
1600, 1715, 2850, 2930, 3055 cm–1. 1H NMR (CDCl3, 500 MHz):
δ = 1.06 (d, J = 7.4 Hz, 3 H), 1.28–1.31 (m, 1 H), 1.37 (qd, J =
11.6, 3.5 Hz, 1 H), 1.51 (tt, J = 12.6, 3.5 Hz, 1 H), 1.58–1.70 (m, 3
H), 2.54 (dt, J = 9.7, 4.7 Hz, 1 H), 2.59–2.65 (m, 2 H), 2.67–2.75
(m, 1 H), 3.50 (ddd, J = 12.9, 9.0, 4.6 Hz, 1 H), 3.66 (dt, J = 12.9,
5.7 Hz, 1 H), 4.11–4.15 (m, 1 H), 6.91 (t, J = 7.0 Hz, 1 H), 7.00 (d,
J = 8.0 Hz, 2 H), 7.32 (t, J = 8.0 Hz, 2 H) ppm. 13C NMR (CDCl3,
125 MHz): δ = 18.08, 20.17, 26.78, 27.66, 28.18, 40.13, 44.27, 55.67,
56.79, 116.66, 119.73, 129.34, 149.60, 209.94 ppm. HRMS
(C16H21NO [M+]): calcd. for 243.1623; found 243.1630. C16H21NO:
calcd. C 78.97, H 8.70, N 5.76; found C 78.81, H 8.62, N 5.51.
4-(8-Phenyl-1,4-dioxa-8-azaspiro[4.5]dec-7-yl)-butanal (8): Dibal-H
(1 m in toluene; 3.5 mL, 3.5 mmol) was added by syringe to a
stirred solution of 7 (1.0 g, 3.5 mmol) in dichloromethane (15 mL),
under an atmosphere of argon, at –60 °C. The temperature was
then raised to –20 °C. After standing overnight at that temperature,
the reaction was quenched by the addition of 20 mL of a 5% HCl
solution. The aqueous phase was made basic by the addition of
an excess of solid NaHCO3, and extracted with dichloromethane
(30 mL×3). The organics were dried with MgSO4 and concentrated
in vacuo to give a slightly yellow oil, which upon purification by
column chromatography (petroleum ether/diethyl ether, 20:80)
yielded 0.78 g (77%) of 8. Rf (petroleum ether/diethyl ether, 20:80)
4-Methyl-NЈ-[(4E,4aS*,5R*,8aR*)-5-methyl-1-phenyloctahydroqui-
nolin-4(1H)-ylidene]benzenesulfonohydrazide (12): Octahydroquino-
linone 10 (1.32 g, 5.43 mmol) was dissolved in 20 mL of EtOH and
refluxed overnight in the presence of p-toluenesulfonohydrazide
(1.11 g, 5.96 mmol). The solvent was evaporated in vacuo, and the
= 0.53. IR (KBr): ν = 703, 737, 1040, 1265, 1499, 1597, 1722, 2958,
˜
1
3054 cm–1. H NMR (CDCl3, 300 MHz): δ = 1.50–1.70 (m, 4 H), resulting crude reaction mixture was taken up in a MeOH/H2O
1.75–1.85 (m, 3 H), 1.97 (dd, J = 14.3, 5.4 Hz, 1 H), 2.37 (t, J =
4.80 Hz, 2 H), 3.15–3.22 (m, 1 H), 3.38 (dt, J = 13.0, 3.4 Hz, 1 H),
3.79 (q, J = 4.70 Hz, 1 H), 3.93–4.04 (m, 4 H), 6.86 (t, J = 7.3 Hz,
1 H), 6.96 (d, J = 7.9 Hz, 2 H), 7.26 (t, J = 7.4 Hz, 2 H), 9.69 (t,
J = 1.6 Hz, 1 H) ppm. 13C NMR (CDCl3, 75 MHz): δ = 19.24,
29.28, 34.16, 36.81, 43.76, 43.81, 56.06, 63.87, 64.62, 107.55,
118.02, 119.75, 129.20, 150.53, 202.52 ppm. HRMS (C17H23NO3
[M+]): calcd. for 289.1678; found 289.1683. C17H23NO3: calcd. C
70.56, H 8.01, N 4.84; found C 70.94, H 8.14, N 4.53.
(80:20) mixture. The crystalline product was filtered off to afford
12 as a white solid (1.86 g, 83%), m.p. 182–184 °C. Rf (petroleum
ether/diethyl ether, 60:40) = 0.18. IR (KBr): ν = 700, 770, 1600,
˜
1650, 2795, 2950, 3220 cm–1. 1H NMR (CDCl3, 300 MHz): δ =
0.82 (d, J = 6.7 Hz, 3 H), 0.98–1.10 (m, 1 H), 1.20–1.60 (m, 5 H),
2.35–2.63 (m, 7 H), 3.10–3.34 (m, 2 H), 3.60–3.70 (m, 1 H), 6.85–
6.98 (m, 3 H), 7.25–7.40 (m, 4 H), 7.90 (d, J = 8.0 Hz, 2 H), 8.00
(br. s, 1 H) ppm. 13C NMR (CDCl3, 75 MHz): δ = 18.74, 20.26,
21.62, 26.06, 27.94, 28.37, 28.96, 45.51, 49.26, 56.39, 117.78,
120.29, 128.20, 129.21, 129.52, 135.36, 144.03, 150.04, 159.65 ppm.
HRMS (C23H29N3O2S [M+]): calcd. for 411.1981; found 411.1983.
1-Phenyl-2,3,6,7,8,8a-hexahydroquinolin-4(1H)-one
(9):
THF
(5 mL) containing 8 (0.76 g, 2.63 mmol) and 50 mL of 1.5 m H2SO4
were added, under an argon atmosphere, to a flask equipped with
a reflux condenser. The solution was refluxed under an argon at-
mosphere for 5 h. The reaction mixture was made basic by the ad-
dition of 50 mL of a saturated solution of Na2CO3. The aqueous
phase was transferred to a separating funnel and extracted with
diethyl ether (30 mL×3). The combined organic layers were dried
with MgSO4 and concentrated. The residue was purified by column
(4aR*,5R*,8aR*)-5-Methyl-1-phenyl-1,2,4a,5,6,7,8,8a-octahydro-
quinoline (13): Compound 12 (1.82 g, 4.43 mmol) in 30 mL of THF
was added with a syringe to an oven-dried Schlenk tube cooled
to –60 °C and filled with argon. Then, nBuLi (1.6 m in hexane;
6.09 mL, 9.74 mmol) was added. After stirring for 90 min at
–60 °C, the solution was allowed to warm to room temperature and
was stirred for 2 h. Then, it was quenched with an excess of water
Eur. J. Org. Chem. 2005, 2269–2280
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2277

N. Girard, J.-P. Hurvois, C. Moinet, L. Toupet
FULL PAPER
(25 mL) and extracted once more with diethyl ether (50 mL×3).
The combined ethereal layers were dried with MgSO4 and concen-
trated. The crude mixture was purified by column chromatography
(petroleum ether/diethyl ether, 70:30) to afford 13 (0.82 g) in 82%
mopropane (0.36 mL, 0.49 g, 3.95 mmol) was added, and the solu-
tion was placed at –20 °C overnight. Then, 15 mL of water was
added, and the resulting solution was extracted with diethyl ether
(50 mL×3). The combined extracts were dried (MgSO4) and con-
centrated to yield a crude orange oil, which was purified by column
yield. R (petroleum ether/diethyl ether, 70:30) = 0.72. IR (KBr): ν
˜
f
= 705, 740, 1265, 1600, 2850, 2930, 3055 cm–1. H NMR (CDCl3, chromatography (petroleum ether/diethyl ether; 95:5) to afford 16
300 MHz): δ = 1.25 (d, J = 7.3 Hz, 3 H), 1.35–1.40 (m, 1 H), 1.55– (0.46 g, 72%) as a viscous oil that solidified upon cooling. A fur-
1.60 (m, 5 H), 2.05–2.17 (m, 1 H), 2.55–2.62 (m, 1 H), 3.57 (dm, J ther slow crystallization from a petroleum ether/diethyl ether (50:1)
1
= 17.4 Hz, 1 H), 3.90 (dq, J = 17.4, 3.2 Hz, 1 H), 4.12–4.23 (m, 1
H), 5.70 (dm, J = 10.1, 3.0 Hz, 1 H), 5.88 (dq, J = 10.1, 3.0 Hz, 1
H), 6.83 (t, J = 7.2 Hz, 1 H), 6.95 (d, J = 8.3 Hz, 2 H), 7.33 (t, J
= 7.0 Hz, 2 H) ppm. 13C NMR (CDCl3, 75 MHz): δ = 18.63, 19.62,
22.07, 27.02, 34.50, 41.45, 44.24, 50.01, 114.00, 117.46, 124.15,
mixture afforded colorless crystals, which were analyzed by X-ray
diffraction; m.p. 100–102 °C. Rf (petroleum ether/diethyl ether,
95:5) = 0.5 ppm. IR (KBr): ν = 700, 750, 1594, 2216, 2871,
˜
1
2923 cm–1. H NMR (CDCl3, 300 MHz): δ = 0.77 (t, J = 6.9 Hz,
3 H), 0.90 (d, J = 6.6 Hz, 3 H), 1.22–1.55 (m, 10 H), 1.71 (dm, J
129.24, 130.27, 149.42 ppm. HRMS (C16H21N [M+]): calcd. for = 14.0 Hz, 1 H), 1.77–2.00 (m, 3 H), 2.02–2.20 (m, 2 H), 3.55 (q,
227.1674; found 227.1669.
J = 2.6 Hz, 1 H), 7.00–7.70 (br. m, 5 H) ppm. 13C NMR (CDCl3,
75 MHz): δ = 14.10, 17.21, 19.67, 20.55, 24.10, 27.61, 30.54, 30.35,
35.72, 41.80, 42.96, 57.45, 63.44, 116.53, 120.98, 126.81, 128.81
(br), 145.55 ppm. HRMS (C20H28N2 [M+]): calcd. for 296.2252;
found 296.2257.
(4aS*,5R*,8aR*)-5-Methyl-1-phenyldecahydroquinoline (14): Com-
pound 13 (0.79 g, 3.48 mmol) was dissolved in 30 mL of methanol
and was placed in a low-pressure hydrogenation apparatus in the
presence of 0.080 g of Pd/C (10%). Stirring for 10 h under one
atmosphere of H2, and subsequent work-up, afforded a crude oil,
which was purified by column chromatography (petroleum ether/
diethyl ether, 98:2), to yield 14 (0.68 g, 85%) as a colorless oil. Rf
(2S*,4aS*,5R*,8aR*)-5-Methyl-1-phenyl-2-propyldecahydroquino-
line (17): NaBH4 (0.23 g, 6.08 mmol) was added to an ethanol solu-
tion (10 mL) of aminonitrile 16 (0.46 g, 1.55 mmol). The solution
was stirred at room temperature overnight and the solvent was then
evaporated. The crude material was taken-up with water (50 mL)
and extracted with diethyl ether (50 mL×3). The organics were
dried (MgSO4) and concentrated to give an oily residue, which was
purified by a rapid filtration through a chromatography column
(petroleum ether) to yield 17 (0.34 g, 81%) as a colorless oil. Upon
cooling, the oily product crystallized as colorless plates (m.p. Ͻ
40 °C) which were analyzed by X-ray diffraction. Rf (petroleum
(petroleum ether/diethyl ether, 98:2) = 0.29. IR (KBr): ν = 695, 740,
˜
1265, 1600, 2850, 2985, 3055 cm–1. 1H NMR (CDCl3, 300 MHz): δ
= 1.16 (d, J = 7.25 Hz, 3 H), 1.21–1.31 (m, 1 H), 1.32–1.42 (m, 1
H), 1.43–1.56 (m, 3 H), 1.58–1.75 (m, 3 H), 1.77–1.92 (m, 4 H),
2.96 (td, J = 12.0, 3.0 Hz, 1 H), 3.34 (dt, J = 12.0, 3.4 Hz, 1 H),
4.00 (dt, J = 11.3, 3.4 Hz, 1 H), 6.78 (t, J = 7.2 Hz, 1 H), 6.97 (d,
J = 8.8 Hz, 2 H), 7.26 (t, J = 7.3 Hz, 2 H) ppm. 13C NMR (CDCl3,
75 MHz): δ = 19.79, 20.93, 21.75, 26.28, 26.36, 27.88, 34.60, 42.47,
43.87, 54.77, 117.02, 118.88, 129.59, 151.61 ppm. HRMS (C16H23N
[M+]): calcd. for 229.1831; found 229.1828.
ether) = 0.42. IR (KBr): ν = 700, 745, 1595, 2795, 2860, 2930,
˜
2955 cm–1. 1H NMR (C6D6, 500 MHz, 343 K): δ = 0.82 (t, J =
7.0 Hz, 3 H), 1.03 (d, J = 6.8 Hz, 3 H), 1.10–1.52 (m, 10 H), 1.55–
1.70 (m, 2 H), 1.75–1.90 (m, 2 H), 2.00–2.10 (m, 1 H), 2.20–2.29
(m, 1 H), 2.98–3.03 (m, 1 H), 3.27–3.32 (m, 1 H), 7.02–7.07 (m, 1
H), 7.17–7.22 (m, 2 H), 7.25–7.31 (m, 2 H) ppm. 13C NMR (C6D6,
125 MHz, 343 K): δ = 14.9, 19.08, 19.47, 20.83, 25.97, 27.08, 29.57,
30.02, 34.25, 37.01, 43.98, 59.44, 60.89, 123.17, 124.36, 128.77,
150.80 ppm. HRMS (C19H29N [M+]): calcd. for 271.2300; found
271.2290. C19H29N: calcd. C 84.07, H 10.77, N 5.16; found C 83.86,
H 10.80, N 5.23.
5-Methyl-1-phenyldecahydroquinoline-2-carbonitrile (15): Decahy-
droquinoline 14 (0.64 g, 2.79 mmol) was dissolved in 250 mL of
methanol containing 5 g of lithium acetate and 0.82 g (16.73 mmol)
of NaCN. The solution was transferred into a divided cell and oxid-
ized at a planar vitreous carbon electrode at +0.80 V vs. SCE. After
the consumption of 660 Coulomb, the electrolysis was stopped and
the solution was concentrated under reduced pressure. The crude
material was taken-up with water (50 mL) and extracted with di-
ethyl ether (50 mL×3). The organics were dried with MgSO4 and
concentrated in vacuo. The crude material was purified by column
chromatography (petroleum ether/diethyl ether, 95:5) to afford 15
(0.55 g, 78%) as a cis/trans mixture (40:60). Rf (petroleum ether/
(2S*,4aS*,5R*,8aR*)-5-Methyl-2-propyldecahydroquinoline
[( )-
Pumiliotoxin C] (1): Absolute EtOH (8 mL), 17 (0.18 g,
0.66 mmol), and 20 mL of liquid ammonia were added, in that or-
der, to a Schlenk tube containing 10 mL of THF at –40 °C. Lithium
wire (0.91 g, 130 mmol) was then added in small pieces over a 1 h
period, upon which the solution became blue. Stirring was contin-
ued for 1 h, and the solution was quenched with an excess of water
(100 mL). The ammonia was allowed to evaporate, and the re-
sulting crude mixture was extracted with diethyl ether (100 mL×2).
The organics were dried with MgSO4 and concentrated to afford a
mixture of crude enamines, which was dissolved in an H2SO4/H2O/
EtOH (0.5:5.0:4.5) mixture. The solution was heated at 60 °C for
10 min and the solvents were then evaporated. The hydrochloride
1 was dissolved in water (10 mL), and the solution was made basic
(pH 12) by addition of NaOH pellets. The free amine 1 was ex-
tracted with diethyl ether (30 mL×2) and the aqueous layer was
discarded. The combined organic layers were dried (MgSO4) and
concentrated to afford crude 1, which was purified by column
chromatography with diethyl ether saturated with gaseous ammo-
nia to give 1 (0.10 g, 78%) as a colorless oil whose NMR spectro-
scopic data were in agreement with those reported in the litera-
ture.[8] Rf (diethyl ether saturated with gaseous ammonia) = 0.60
diethyl ether) = 0.40. IR (KBr): ν = 699, 769, 1491, 1596, 2222,
˜
1
2228, 2865, 2935 cm–1. H NMR (CDCl3, 200 MHz): δ = 0.91 (d,
J = 6.7 Hz, 3 H), 1.11 (d, J = 7.2 Hz, 3 H), 1.24–2.24 (m, 24 H),
3.55 (q, J = 3.0 Hz, 1 H), 3.95–4.08 (m, 1 H), 4.15–4.18 (m, 1 H),
4.40 (br. s, 1 H), 6.94–7.40 (m, 10 H) ppm. 13C NMR (CDCl3,
50 MHz): δ = 19.36, 19.70, 20.37, 20.44, 23.03, 23.53, 24.77, 27.80,
29.34, 29.48, 35.40, 42.04, 42.65, 54.22, 57.73, 118.20, 121.06,
125.85, 126.74, 129.13, 129.50, 147.73, 148.51 ppm. HRMS
(C17H22N2 [M+]): calcd. for 254.1783; found 254.1790. C17H22N2:
calcd. C 80.27, H 8.72, N 11.01; found C 80.02, H 8.79, N 10.39.
(2S*,4aS*,5R*,8aR*)-5-Methyl-1-phenyl-2-propyldecahydroquino-
line-2-carbonitrile (16): Aminonitrile 15 (0.55 g, 2.16 mmol, cis/
trans, 40:60) was dissolved in 10 mL of THF in a Schlenk tube
equipped with a rubber stopper and flushed with argon. The re-
sulting solution was cooled to –60 °C and a solution of LDA [pre-
pared from 1.6 m nBuLi in hexane (2.05 mL, 3.28 mmol) and diiso-
propylamine (0.52 mL, 0.37 g, 3.71 mmol)] in THF was added with
a syringe. The anion solution was stirred at –60 °C for 1 h, 1-bro-
2278
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2269–2280

Total Synthesis of ( )-Pumiliotoxin C
FULL PAPER
Table 2. X-ray crystallographic data for 16 and 17.
16
17
Formula
Mol. mass
C20H28N2
296.44
C19H29N
271.43
Cryst. syst.
triclinic
P1
1.146
triclinic
P1
1.111
¯
¯
Space group
Dx [Mgm–3]
a [Å]
b [Å]
c [Å]
α [°]
6.4442(1)
10.6710(2)
12.9573(2)
97.131(1)
91.583(1)
102.251(1)
862.28(2)
2
5.535(4)
8.364(4)
17.843(9)
84.47(4)
83.62(5)
82.62(5)
811.3(8)
2
β [°]
γ [°]
V [Å3]
Z
F(000)
324
0.65
300
0.63
μ [cm–1]
λ(Mo-Kα) [Å]
T [K]
0.71073
120
0.71073
293
Crystal size [mm]
Radiation
0.48×0.34×0.30
Mo-Kα
0.42×0.33×0.33
Mo-Kα
Max. 2θ [°]
Scan
54
54
ω/2Θ = 1
ω/2Θ = 1
hkl range
tmax. [s]
0Ǟ80, –13Ǟ13, –16Ǟ16
15
0Ǟ7, –10Ǟ10, –22Ǟ22
60
Reflections measured
Reflections observed [I Ͼ 2.0σ(I)]
Final R
Rw
7447
3341
0.042
0.140
3906
2906
0.042
0.113
[4]
[5]
A. I. Meyers, G. Milot, J. Am. Chem. Soc. 1993, 115, 6652–
6660.
Z. H. Shao, F. Z. Peng, J. B. Chen, C. Y. Wang, R. Huang,
Y. Q. Tu, L. Li, H. B. Zhang, Synth. Commun. 2004, 34, 2031–
2038.
G. Mehta, M. Praveen, J. Org. Chem. 1995, 60, 279–280.
M. Akashi, Y. Sato, M. Mori, J. Org. Chem. 2001, 66, 7873–
7874.
T. G. Back, K. Nakajima, Tetrahedron Lett. 1997, 38, 989–992.
N. A. LeBel, N. Balasubramanian, J. Am. Chem. Soc. 1989,
111, 3363–3368.
IR (KBr): ν = 755, 1078, 1311, 1449, 2605, 2719, 2796, 2862,
˜
1
2926 cm–1. H NMR (CDCl3, 500 MHz): δ = 0.84 (d, J = 6.6 Hz,
3 H), 0.91 (t, J = 6.9 Hz, 3 H), 0.94 (qd, J = 10.0, 3.3 Hz, 1 H),
1.05–1.15 (m, 2 H), 1.30–1.42 (m, 8 H), 1.45–1.64 (m, 4 H), 1.82–
1.90 (m, 1 H), 1.87–1.98 (dm, J = 11.0 Hz, 1 H), 2.52–2.58 (m, 1
H), 2.82–2.84 (m, 1 H) ppm. 13C NMR (CDCl3, 125 MHz): δ =
14.33, 19.16, 19.92, 21.27, 27.07, 27.38, 29.71, 33.44, 35.95, 39.76,
42.60, 55.97, 57.73 ppm. HRMS (C13H25N [M+]): calcd. for
195.1987; found 195.1982.
[6]
[7]
[8]
[9]
X-ray Crystallographic Study: Crystallographic data were collected
on a Nonius Kappa CCDC (for 16) or on an automatic dif-
fractometer CAD4 Nonius (for 17) with graphite-monochromated
Mo-Kα radiation. Details are given in Table 2. The structures were
solved with SIR-97,[37] which revealed the non-hydrogen atoms of
the molecules. Refinement was performed by full-matrix least-
squares techniques with SHELXL-97.[38] Atom scattering factors
were taken from the International Tables for X-ray Crystallogra-
phy.[39] Figures were drawn with PLATON98[40] or ORTEP-3 for
Windows.[41] CCDC-221915 (16) and –226074 (17) contain the sup-
plementary crystallographic data for this paper. These data can be
obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
[10]
[11]
[12]
L. E. Overman, P. J. Jessup, J. Am. Chem. Soc. 1978, 100, 5179–
5185.
W. Oppolzer, C. Fehr, J. Warneke, Helv. Chim. Acta 1977, 60,
48–58.
K. Maruoka, T. Miyazaki, M. Ando, Y. Matsumura, S. Sakane,
K. Hattori, H. Yamamoto, J. Am. Chem. Soc. 1983, 105, 2831–
2843.
R. P. Polniaszek, L. W. Dillard, J. Org. Chem. 1992, 57, 4103–
4110.
D. L. Comins, A. Dehghani, Tetrahedron Lett. 1991, 32, 5697–
5700.
M. Bonin, R. Besselièvre, D. S. Grierson, H. P. Husson, Tetra-
hedron Lett. 1983, 24, 1493–1496.
E. W. Dijk, L. Panella, P. Pinho, R. Naasz, A. Meetsma, A. J.
Minnaard, B. L. Feringa, Tetrahedron 2004, 60, 9687–9693.
W. Oppolzer, E. Flaskamp, L. W. Bieber, Helv. Chim. Acta
2001, 84, 141–145.
[13]
[14]
[15]
[16]
[17]
[18]
[19]
Acknowledgments
M. Weymann, M. Schultz-Kukula, H. Kunz, Tetrahedron Lett.
1998, 39, 7835–7838.
N. G. would like to thank the MENRT for a grant.
T. Riechers, H. C. Krebs, R. Wartchow, G. Habermehl, Eur. J.
Org. Chem. 1998, 2641–2646.
T. G. Back, K. Nakajima, J. Org. Chem. 1998, 63, 6566–6571.
N. Girard, C. Gautier, R. Malassene, J. P. Hurvois, C. Moinet,
L. Toupet, Synlett 2004, 2005–2009.
[1] T. F. Spande, P. Jain, H. M. Garraffo, L. K. Pannell, H. J. C.
Yeh, J. W. Daly, J. Nat. Prod. 1999, 62, 5–21.
[2] J. W. Daly, B. Witkop, T. Tokuyama, T. Nishikawa, I. L. Karle,
Helv. Chim. Acta 1977, 60, 1128–1140.
[3] J. W. Daly, T. Tokuyama, G. Habermehl, I. L. Karle, B. Wit-
kop, Justus Liebigs Ann. Chem. 1969, 729, 198–204.
[20]
[21]
[22]
P. Caubère, B. Loubinoux, Bull. Soc. Chim. Fr. 1968, 3857–
3861.
Eur. J. Org. Chem. 2005, 2269–2280
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2279

N. Girard, J.-P. Hurvois, C. Moinet, L. Toupet
FULL PAPER
[23] E. Le Gall, J. P. Hurvois, T. Renaud, C. Moinet, A. Tallec, P.
Uriac, S. Sinbandhit, L. Toupet, Liebigs Ann./Recueil 1997,
2089–2101.
[33] For a similar regioselectivity in the anodic oxidation of piperi-
dines see: T. Shono, Y. Matsumura, K. Tsubata, Y. Sugihara,
S. Yamane, T. Kanazawa, T. Aoki, J. Am. Chem. Soc. 1982,
104, 6697–6703.
[34] T. Shono, Y. Matsumara, O. Onomura, Y. Yamada, Tetrahe-
dron Lett. 1987, 28, 4073–4074.
[35] Suggested by a referee: This preferred conformation in α-ami-
nonitriles has been likened to the anomeric effect.
[36] E. Le Gall, J. P. Hurvois, S. Sinbandhit, Eur. J. Org. Chem.
1999, 2645–2653.
[37] A. Altomare, M. C. Burla, M. Camalli, G. Cascarano, C. Gia-
covazzo, A. Guargliardi, A. G. G. Moliterni, G. Polidori, R.
Spagna, J. Appl. Crystallogr. 1999, 32, 115–119.
[38] G. M. Sheldrick, SHELX-97, University of Göttingen, Ger-
many, 1997.
[39] International Tables for X-ray Crystallography, 1992, vol. C,
Kluwer Academic Publishers, Dordrecht.
[40] A. L. Spek, PLATON, A Multipurpose Crystallographic Tool,
Utrecht University, Utrecht, The Netherlands, 1998.
[41] L. J. Farrugia, J. Appl. Crystallogr. 1998, 30, 565–567.
Received: November 29, 2004
[24] For the alkylation of α-aminonitriles derived from related pip-
eridinic systems, see: T. Shono, J. Terauchi, K. Kitayama, Y.
Takeshima, Y. Matsumura, Tetrahedron 1992, 48, 8258–8262.
[25] For a similar NOE effect in the decahydroquinoline series, see:
T. Tokuyama, T. Tsujita, A. Shimada, H. M. Garraffo, T. F.
Spande, J. W. Daly, Tetrahedron 1991, 47, 5401–5414.
[26] F. W. Vierhapper, E. L. Eliel, J. Org. Chem. 1977, 42, 51–62.
[27] H. Booth, D. V. Griffiths, J. Chem. Soc., Perkin Trans. 2 1975,
111–118.
[28] P. Deslongchamps, in Stereoelectronic Effects in Organic Chem-
istry, Pergamon, New York, 1983, chapter 6.
[29] For the sake of clarity, alternate pathways b and bЈ were not
considered as they lead to unfavorable boatlike transition
states. For a comprehensive review see : R. V. Stevens, Acc.
Chem. Res. 1984, 17, 289–296.
[30] For a review see: R. H. Shapiro, in Organic Reactions, John
Wiley & Sons, New York, 1976, vol. 23, pp. 405–507.
[31] R. M. Adlington, A. G. M. Barrett, Acc. Chem. Res. 1983, 16,
55–59.
[32] R. H. Shapiro, M. F. Lipton, K. J. Kolonko, R. L. Buswell,
L. A. Capuano, Tetrahedron Lett. 1975, 16, 1811–1814.
2280
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2269–2280