A gram-scale batch and flow total synthesis of perhydrohistrionicotoxin

Article abstract of DOI:10.1002/chem.201001435

The total synthesis of the spi- ropiperidine alkaloid (-)-perhydrohis- trionicotoxin (perhydro-HTX) 2 has been accomplished on a gram scale by employing both conventional batch chemistry as well as microreactor techniques. )-6-Pentyltetrahydro- pyran-2-one 8 underwent nucleophilic ring opening to afford the alcohol 10, which was elaborated to the nitrone 13. Protection of the nitrone as the 1,3- adduct of styrene and side-chain extension to the unsaturated nitrile afforded a precursor 17, which underwent dipolar cycloreversion and 1,3-dipolar cycloaddition to give the core spirocyclic precursor 18 that was converted into perhydro-HTX 2. The principal steps to the spirocycle 18 have successfully been transferred into flow mode by using different types of microreactors and in a telescoped fashion, allowing for a more rapid access to the histrioni- cotoxins and their analogues by continuous processing.

Full text of DOI:10.1002/chem.201001435

FULL PAPER
DOI: 10.1002/chem.201001435
A Gram-Scale Batch and Flow Total Synthesis of Perhydrohistrionicotoxin
Malte Brasholz,[a] James M. Macdonald,[a] Simon Saubern,[a] John H. Ryan,*[a] and
Andrew B. Holmes[a, b]
Abstract: The total synthesis of the spi-
ropiperidine alkaloid (À)-perhydrohis-
Protection of the nitrone as the 1,3-
adduct of styrene and side-chain exten-
sion to the unsaturated nitrile afforded
a precursor 17, which underwent dipo-
lar cycloreversion and 1,3-dipolar cy-
cloaddition to give the core spirocyclic
precursor 18 that was converted into
perhydro-HTX 2. The principal steps
to the spirocycle 18 have successfully
been transferred into flow mode by
using different types of microreactors
and in a telescoped fashion, allowing
for a more rapid access to the histrioni-
cotoxins and their analogues by contin-
uous processing.
trionicotoxin (perhydro-HTX)
2 has
been accomplished on a gram scale by
employing both conventional batch
chemistry as well as microreactor tech-
niques.
(S)-(À)-6-Pentyltetrahydro-
Keywords: alkaloids
·
cycloaddi-
pyran-2-one 8 underwent nucleophilic
ring opening to afford the alcohol 10,
which was elaborated to the nitrone 13.
tion · flow chemistry · olefination ·
total synthesis
Introduction
which has been achieved by a combination of conventional
batch techniques and microreactor technology. The histrioni-
cotoxins[4] constitute a class of poison-arrow frog alkaloids
that have gained the considerable attention of chemists and
biologists over almost 40 years; numerous approaches to
these structurally challenging target molecules have been
developed in the past.[5] We have contributed to this area by
the use of intramolecular nitrone dipolar cycloaddition strat-
egies and the continued improvement of practical proce-
dures, leading to several general and highly competitive dia-
stereoselective and enantioselective syntheses[6] of the
family, including (À)-histrionicotoxin 285A (1) and 2.
The recognition of microreactor technology as a competitive
way of processing chemical intermediates has been an im-
portant development in recent years, and synthesis under
continuous flow conditions in many cases offers distinct ad-
vantages over conventional round-bottomed flask chemis-
try.[1] Microreactors enable the precise control of reaction
parameters, linear scaleup and safe handling of hazardous
substances with exquisite reproducibility. Many key chemi-
cal transformations transfer readily from batch into flow
chemical procedures,[2] and flow chemistry can be a substan-
tial support for diversity-oriented synthesis programs, and
even has significant applications in the total synthesis of nat-
ural products and drug molecules.[3]
Herein, we describe the total synthesis of the reduced spi-
ropiperidine alkaloid (À)-perhydrohistrionicotoxin (2),
[a] Dr. M. Brasholz, Dr. J. M. Macdonald, Dr. S. Saubern,
Dr. J. H. Ryan, Prof. Dr. A. B. Holmes
CSIRO Molecular and Health Technologies, Bayview Avenue
Clayton, VIC 3168 (Australia)
Fax : (+61)3-9545-2446
E-mail: jack.ryan@csiro.au
Our ongoing interest in the histrionicotoxins is fuelled by
their biological activity, such as serving as non-competitive
inhibitors of the neuromuscular, ganglionic and nicotinic
acetylcholine receptors.[7] However, the full medicinal poten-
tial of the histrionicotoxins, or simplified analogues thereof,
remains to be explored. From a synthetic point of view, two
nitrone-based dipolar cycloaddition approaches appear to
[b] Prof. Dr. A. B. Holmes
Prof. Dr. A. B. Holmes, Bio 21 Institute, School of Chemistry
University of Melbourne, Flemington Road
Parkville, VIC 3010 (Australia)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201001435.
Chem. Eur. J. 2010, 16, 11471 – 11480
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
11471

be the most suitable for the rapid access to diverse natural
and non-natural members of the compound class
(Scheme 1): Firstly, the domino cyclisation of bis-nitrile
account describes a gram-scale batch synthesis of 2, and
demonstrates translation to continuous flow conditions.
Results and Discussion
For the gram-scale synthesis of 2, we required a substantial
quantity of the n-pentyl-substituted 6,6,5-tricyclic isoxazoli-
dine intermediate 18 (Scheme 2). While our previous ap-
proach to a related hydroxymethyl derivative involved a
lengthy elaboration of (S)-(À)-glycidol,[6e] we consequently
employed (S)-(À)-6-pentyltetrahydropyran-2-one (8) as the
chiral starting material.[10] Nucleophilic addition of the lithi-
um acetylide derived from alkyne 9[11] to the lactone 8 at
À788C provided the propargyl ketone product 10 in near
quantitative yield. The alkyne 10 was hydrogenated over
palladium(II) hydroxide on carbon in the presence of trie-
thylamine to afford the saturated derivative 11. Sulfonyla-
tion of the alcohol gave the mesylate 12, which was convert-
ed into the nitrone 13 by formation of the oxime using an
excess of hydroxylammonium chloride and sodium bicar-
bonate at 708C in ethanol, followed by intramolecular nu-
cleophilic displacement of the sulfonate. Nitrone 13 was dif-
ficult to obtain in pure form, and therefore, it was immedi-
ately processed further by protection as its styrene adduct
14. The 1,3-dipolar cycloaddition of nitrone 13 with styrene
proceeded with useful diastereoselectivity to provide the
separable exo isomer of isoxazolidine 14 in 44% over three
steps from alcohol 11. Acidic hydrolysis of acetal 14 led to
aldehyde 15 and Peterson olefination of this intermediate by
using Kojimaꢁs reagent 16[12] gave a,b-unsaturated nitrile 17
Scheme 1. Retrosynthetic analysis of spirocycles 3 and 5.
ketone 4 with hydroxylamine, which elegantly leads to the
racemic bifunctional 6,6,5-tricyclic bis-nitrile 3,[8,6d] and sec-
ondly, the concise route to enantiomerically pure azaspirocy-
cle 5 via cyclic nitrone 6, which derives from a chiral d-lac-
tone 7.[6e] The tricyclic isoxazolidines 3 and 5 are late-stage
precursors to many members of the compound class of the
histrionicotoxins, and they can also be regarded as ideal
templates for the synthesis of novel analogues.
We have recently reported an improved batch synthesis of
the ketone 4 and its conversion into isoxazolidine 3 under
continuous flow conditions.[9] We further aimed at investigat-
ing a more generic route from chiral lactones such as 7 for
flow processing based on a novel batch synthesis of 2. This
Scheme 2. Batch synthesis of 6,6,5-tricyclic isoxazolidine (À)-18. MsCl=mesyl chloride.
11472
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11471 – 11480

Total Synthesis of Perhydrohistrionicotoxin
FULL PAPER
with excellent Z selectivity (Z/E 95:5). Finally, microwave
heating of nitrile 17 at 1848C led to loss of styrene by 1,3-di-
polar cycloreversion and subsequent intramolecular 1,3-di-
polar cycloaddition to provide the enantiopure 6,6,5-tricyclic
nitrile (À)-18 in 70% yield.
workup and drying in vacuum provided the propargyl
ketone 10 in 83% yield, and in a purity of 85–90%, as esti-
mated by 1H NMR spectroscopy. As an alternative to the
aqueous workup, an Omnifit-cartridge[16] containing Quadra-
Pure-IDA dicarboxylic acid resin[17] could be connected to
the reactor outlet, leading to a quench of the lithium alkox-
ide on the polymeric reagent. A noticeable feature of this
telescoped flow procedure was the surprisingly long reaction
time of around 30 min required for the deprotonation of the
terminal alkyne 9 with n-butyllithium at 08C, whereas the
nucleophilic attack of the resulting lithium acetylide anion
on lactone 8 at the same temperature proceeded to comple-
tion within five minutes or less.
As it stands, the route to isoxazolidine 18 represents an
efficient and competitive means of obtaining this late-stage
intermediate to 2 and its analogues, and it could be prepared
enantiomerically pure in quantities in excess of two grams.
However, we were interested in further simplification of
processing, with particular emphasis on shortening of reac-
tion times and reducing the number of purification steps,
and we envisaged realising this objective using flow-chemis-
try techniques. The first challenge was the use of organome-
tallic reagents under continuous flow conditions, and the
timely addition of reagent volumes in telescoped flow chem-
ical procedures. As shown in Scheme 3, the investigation
started with the nucleophilic addition of lithiated alkyne 9
to racemic d-lactone 8, and after careful investigation of the
reaction parameters, it could be performed in a telescoped
three-channel flow chemical procedure at 08C, as an alter-
native to the batch reaction at À788C. As a process configu-
ration, we used two piston pumps compatible with corrosive
fluids (Ismatec)[13] in combination with a standalone HPLC
pump (Knauer),[14] the mixing of the individual reaction
channels were performed on glass chips (Sigma Aldrich).[15]
A stock solution of the alkyne 9 in tetrahydrofuran was
mixed with a stock solution of n-butyllithium in hexane, on
a first glass chip, and the resulting flow stream was directed
into a reactor coil made out of PTFE tubing (internal
volume 30 mL, 1/8 in. o.d., 1.5 mm i.d.), where complete de-
protonation occurred within a retention time of roughly
30 min. As soon as a constant flow stream of the lithiated
alkyne emerged from the first reactor coil, a stock solution
of the lactone (Æ)-8 in tetrahydrofuran was channelled in
on a second glass chip positioned after the first coil, and the
resulting flow stream was passed through a second reactor
coil (internal volume 10 mL), and into an aqueous solution
of ammonium chloride at the reactor outlet. Extractive
The flow procedure shown in Scheme 3 could also be per-
formed under steady-state conditions, delivering the pro-
AHCTUNGTREGpNNNU argyl ketone product 10 on a gram scale and with sufficient
purity to be used in the next stage. As shown in Scheme 4,
Scheme 4. Flow hydrogenation of alkyne 10 to the saturated alcohol 11.
the crude alkyne 10 was subsequently subjected to flow hy-
drogenation in the H-cube hydrogenator (Thales Nano).[18]
Hydrogenation using palladium(II) hydroxide on carbon as
the catalyst at 408C in ethyl acetate containing Hꢂnigꢁs
base, gave the saturated derivative 11 in near quantitative
yield and in excellent purity. We have performed the hydro-
genation of the alkyne 10 using starting material of varying
purity. Under normal circumstances the alkyne 10 was con-
taminated with small amounts of the precursor alkyne 9,
which had been used in a slight excess in the previous step.
While the alkyne 9 could be removed from the product 11
under high vacuum, it was more practical to hydrogenate
the mixture of the alkynes 10 and 9 because the reduction
product from 9 is more volatile and is easily removed upon
concentration of the product solution under vacuum.
Since minor impurities in the alkyne 10 were unproble-
matic in the hydrogenation step, and the alcohol 11 was typ-
ically isolated in a pure state after workup, we envisaged
combining the telescoped ring opening of lactone 8 and sub-
sequent hydrogenation into a single flow process. To con-
duct experiments on a small scale of typically 0.2 to
0.5 mmol, we built a reactor system consisting of two Va-
pourtec R2+/R4 units;[19] the three reaction channels were
controlled by two R2+ pump modules. The R2+ pump
module enables the controlled small volume dosage of re-
agents through the in-built injection loops and by precise
adjustment of the flow rates of the individual reaction chan-
nels, telescoped plug-flow synthesis involving several flow
channels can be performed in a reliable and reproducible
Scheme 3. Telescoped ring opening of lactone 8 to give propargyl ketone
10.
Chem. Eur. J. 2010, 16, 11471 – 11480
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11473

J. H. Ryan et al.
way. For the timely addition of reagents, we relied on the
calculation of the residence times of the reagent volumes.
As shown in Scheme 5, a solution of the alkyne 9 in tert-
tone 8 was most likely attributed to the competition be-
tween 1,2-addition of the lithiated alkyne to the lactone and
the enolisation of the carbonyl group by the residual diiso-
propylamine. At the same time, the presence of free diiso-
propylamine is beneficial in the terminal hydrogenation
step, where an amine base is required to avoid the forma-
tion of decomposition or over-reduction products during the
reduction of the intermediate alkyne 10 over the palladium
catalyst.[6e]
The two-step flow preparation of alcohol 11 shown in
Schemes 3 and 4 was the preferred method to obtain this
intermediate on a preparative scale, and as shown in
Scheme 6, alcohol 11 was subsequently elaborated into the
Scheme 5. Telescoped flow synthesis of alcohol 11. QP-IDA=Quadra-
Pure-IDA dicarboxylic acid resin.
butyl methyl ether (TBME), and a solution of lithium diiso-
propylamide (LDA) in the same solvent were injected into
sample loops of a Vapourtec R2+ unit. The two reagent
channels were combined at ambient temperature in a Y-
shaped PEEK mixer piece, and the resulting stream was di-
rected into a 5 mL flow coil (PTFE tubing, 1/16 in. o.d.,
1 mm i. d.) which was cooled to 08C. After the calculated
residence time had passed, a solution of the lactone (Æ)-8 in
TBME was channelled in through a second Y piece, and the
resulting stream was passed through a second 5 mL coil at
08C. The flow stream was passed through an Omnifit car-
tridge containing QuadraPure-IDA dicarboxylic acid resin,
before entering the H-cube hydrogenator, where reduction
took place over palladium(II) hydroxide on carbon at 408C.
The flow stream was collected at the reactor outlet, to pro-
vide the saturated alcohol 11 as a mixture with the unreact-
Scheme 6. Elaboration of alcohol 11 to isoxazolidine 14.
isoxazolidine 14 by flow and batch microwave processing. A
stock solution of the alcohol 11 and Hꢂnigꢁs base in di-
chloromethane was mixed on a glass chip with a solution of
MsCl in the same solvent, at ambient temperature. Within
only one minute, conversion of the alcohol was complete
and a non-aqueous workup by filtration over basic alumina
provided the pure mesylate 12 without any residual MsCl as
a contaminant. The mesylate 12 was then converted into sty-
rene adduct 14, via the oxime and nitrone, in a convenient
and straightforward one-pot microwave reaction. The sub-
strate 12 was mixed with an excess of hydroxylammonium
chloride and Hꢂnigꢁs base in neat styrene, and the mixture
was heated in a microwave vial to 1058C for 6 h. By way of
this much simplified one-pot procedure, the reaction time
for the conversion of mesylate 12 into isoxazolidine 14 could
be significantly reduced, and workup of the sensitive inter-
mediate nitrone 13 could be avoided. Inside a 20 mL micro-
wave vial, we could process 1 mmol of substrate at a time,
and we did not further attempt to develop a flow chemical
procedure for this transformation. Isoxazolidine 14 was iso-
lated in a yield of 40%, as a mixture of the exo and endo
isomer (ratio 85:15).[21]
1
ed lactone 8; conversion of the latter was 60% by H NMR
spectroscopy. After chromatographic purification, the prod-
uct 11 was isolated in a yield of 49% (thus 82% based on
conversion of lactone 8).
By the described procedure we could demonstrate that
the intermediate 11 could be produced in a single operation
from starting materials 8 and 9; however, the conversion of
the lactone in the experiment shown in Scheme 5 was not
complete. The use of an organometallic base such as LDA
or n-butyllithium in such a process poses several prob-
lems.[20] When n-butyllithium in hexane was employed as the
base in an analogous experiment, the conversion of the lac-
tone 8 was increased to 80%. However, reactor clogging by
hydrolysis of the base to lithium hydroxide occurred rapidly,
despite all precautions for the exclusion of moisture, making
such a procedure impractical. On the other hand, LDA in
TBME was much easier to handle, and extensive hydrolysis
was avoided, but the diminished conversion (to 60%) of lac-
11474
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11471 – 11480

Total Synthesis of Perhydrohistrionicotoxin
FULL PAPER
Next, we turned our attention to the Peterson olefination
of aldehyde 15 with Kojimaꢁs reagent 16[12] to prepare the
a,b-unsaturated nitrile 17 under flow conditions. When re-
agent 16 was deprotonated with n-butyllithium in a flask,
and the solution of the lithio-anion in tetrahydrofuran was
mixed on a chip with a solution of the aldehyde 15 in the
same solvent at 08C, we observed a complete conversion
into nitrile 17 within just two minutes to give the same Z/E
selectivity as that observed in the batch reaction at À788C.
To telescope the deprotonation of silylacetonitile reagent 16
and the olefination of aldehyde 15, we again used the reac-
tor configuration shown in Scheme 5, and investigated the
process on a small scale. Since we had previously encoun-
tered difficulties using n-butyllithium, we changed the sol-
vent to toluene and used potassium bis-trimethylsilylamide
(KHMDS) as the base. Gratifyingly, these conditions led to
a comparable result and the nitrile 17 was obtained with
hyde 15 in toluene was channelled in through a second Y
piece. After the combined flow stream had passed through a
second reactor coil at 08C, a quench was performed by
using QuadraPure-IDA dicarboxylic acid resin. The crude
product was collected, and charged onto a short plug of
silica gel. This was first flushed with dichloromethane, which
exclusively removed all residual reagent 19 and the siloxane
byproducts. Thereafter, flushing with ethyl acetate gave the
highly pure olefination product 17 in 83% yield, in a Z/E
1
ratio of >90:10 by H NMR spectroscopy. The subsequent
thermal cycloreversion–cycloaddition of nitrile 17 to tricyclic
isoxazolidine 18 proceeded readily under flow conditions, at
1908C in a steel reactor coil, within a reaction time of
20 min, and a final column chromatography provided the
separable (racemic) product (Æ)-18 in 62% yield.
We completed the total synthesis of 2 by batch processing,
within three further steps, as shown in Scheme 8. The final
stages of the conversion of the polycyclic isoxazolidine 18
followed the strategy developed for a related molecule.[6e]
Reduction with diisobutylaluminium hydride and Wittig ole-
fination of the intermediate aldehyde with the ylide derived
from propyltriphenylphosphonium bromide afforded the tri-
1
>90:10 ratio of Z/E isomers (estimated by H NMR spec-
troscopy), demonstrating that neither the counter cation nor
the solvent have a significant effect on the stereochemical
course of the reaction.
In the batch synthesis, we found the purification of a,b-
unsaturated nitrile 17 was quite tedious, because the silox-
ane side products generated in the Peterson reaction as well
as any excess of reagent 16 show a very similar Rf value to
the product 17 on silica gel. To avoid an additional chroma-
tography step in our flow synthesis, we therefore aimed to
modify reagent 16 in a way that would enable more conve-
ACHTUNGTRENNUNGnient removal of the silyl byproducts of the reaction. Conse-
quently, we prepared silylacetonitrile reagent 19, with a
longer alkyl chain, by a procedure analogous to the prepara-
tion of reagent 16,[12] using 2-methylundecan-2-ol[22] instead
of tert-butanol. The telescoped Peterson olefination using
this novel silylacetonitrile reagent is shown in Scheme 7. A
flow stream of reagent 19 in toluene was mixed in a Y piece
with a stock solution of KHMDS in toluene, at ambient
temperature. The resulting flow stream was directed into a
5 mL coil cooled to 08C, and thereafter, a solution of alde-
Scheme 8. Conversion of tricyclic isoxazolidine 18 into (À)-perhydrohis-
trionicotoxin 2.
Scheme 7. Telescoped Peterson olefination of aldehyde 15 with the modified silylacetonitrile reagent 19 and thermal 1,3-dipolar cycloreversion–cycload-
dition of isoxazolidine 17 to 6,6,5-tricyclic isoxazolidine (Æ)-18.
Chem. Eur. J. 2010, 16, 11471 – 11480
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11475

J. H. Ryan et al.
hexanes, with a boiling range from 40–608C). 1H and 13C NMR spectra
were recorded on a Bruker AV400 spectrometer at 400 and 100.6 MHz,
cyclic alkene 20, which underwent concomitant N,O-bond
cleavage and double-bond reduction upon hydrogenation
over palladium(II) hydroxide on carbon. Enantiomerically
pure 2[23] was obtained from the batch synthesis in a quantity
of one gram, and the ease of its preparation from isoxazoli-
dine 18 implicates once more the potential utility of this
template for the rapid production of analogues.
respectively, or on
a Bruker DRX500 spectrometer at 500 and
125.8 MHz, respectively, using solutions in CDCl3. Chemical shifts are re-
ported relative to the resonances of CHCl3 at d=7.25 ppm (H) and d=
77.0 ppm (C). Microanalyses were performed by Campbell Microanalyti-
cal Laboratory, University of Otago, Dunedin, New Zealand. Melting
points were recorded on an Electrothermal IA9300 digital melting point
apparatus and are uncorrected. Positive-ion EI mass spectra were run on
a ThermoQuest MAT95XL mass spectrometer using an ionisation energy
of 70 eV. Accurate mass measurements were obtained with a resolution
of 5000–10000 using perfluorokerosene as the reference compound.
High-resolution positive-ion electrospray mass spectra were acquired
with a Micromass Q-TOF II mass spectrometer using a cone voltage of
50 V and a capillary voltage of 3.0 kV. The sample was introduced by
direct infusion at a rate of 5 mLminÀ1 using PEG400 as an internal cali-
brant. Unless otherwise indicated, flash chromatography was carried out
by using Merck Kieselgel 60 (230–400 mesh; particle size 0.04–0.63 mm)
silica gel. “Silica gel funnel chromatography” was performed in an analo-
gous fashion as flash chromatography except that a sintered glass funnel
was employed (silica bed: width 8 cm, height 7 cm) and passage of sol-
vents was accelerated through the silica bed by reduced rather than posi-
tive pressure. Analytical TLC was conducted on Sigma–Aldrich silica gel
coated aluminium sheets and visualised with UV and/or by phosphomo-
lybdic acid/EtOH or potassium permanganate reagents.
Conclusion
The overall sequence starting from an enantiopure d-lactone
8 and the alkyne 9 leading to the 6,6,5-tricyclic isoxazolidine
18, and subsequently, to the enantiopure perhydrohistrioni-
cotoxin spiropiperidine structure represents one of the most
efficient means of synthesising these challenging and biolog-
ically valuable target molecules today. We have demonstrat-
ed the effective use of microreactors and have succeeded in
supporting the synthesis program by way of a multi-step
flow synthesis of key building block isoxazolidine 18, in ra-
ACHTUNGTRENNUNG
Batch synthesis of 6,6,5-tricyclic isoxazolidine (À)-18
(9S)-1,1-Diethoxy-9-hydroxy-tetradec-3-yne-5-one 10: A solution of 4,4-
diethoxybut-1-yne 9[11] (12.1 g, 85.4 mmol) in THF (250 mL) was cooled
to À788C. nBuLi (53.7 mL, 1.59m in hexane, 85.4 mmol) was added drop-
wise (10 min) and the pale yellow solution was stirred (30 min). A solu-
tion of the lactone (S)-(À)-8[10] (12.1 g, 71.2 mmol) in THF (100 mL) was
added dropwise (15 min) and this solution was stirred for a further 1 h. A
saturated aqueous solution of NH4Cl (100 mL) was added and the mix-
ture was allowed to warm to RT. This mixture was diluted with H2O and
EtOAc and the organic layer was separated. The aqueous layer was ex-
tracted (2ꢃ) with EtOAc and the combined organic layers were washed
with saturated aqueous solutions of NaHCO3 and NaCl, dried (MgSO4),
filtered and concentrated. The residue was purified by silica gel funnel
chromatography (EtOAc/hexanes containing 1% Et3N, gradient; 10:90,
15:85, 20:80 then 25:75) to yield 10 as a clear, colourless oil (21.6 g,
97%). Rf 0.39 (25:75 EtOAc/toluene); [a]2D5 =À1.5 (c 1.13, CHCl3);
1H NMR (CDCl3, 400 MHz): d=0.86 (t, J=6.9 Hz, 3H), 1.20 (t, J=
7.1 Hz, 3H), 1.23–1.51 (m, 10H), 1.66–1.88 (m, 2H), 2.55 (t, J=7.2 Hz,
2H), 2.68 (d, J=5.6 Hz, 2H), 3.49–3.59 (m, 3H), 3.61–3.71 (m, 2H),
4.68 ppm (t, J=5.6 Hz, 1H); 13C NMR (CDCl3, 100 MHz): d=14.0, 15.1,
20.0, 22.6, 25.2, 25.4, 31.8, 36.5, 37.4, 45.2, 62.1, 71.3, 81.7, 89.2, 100.0,
187.9 ppm; IR (neat) 3473, 2929, 2868, 2218, 1710, 1675, 1455, 1371, 1343,
1227, 1119, 1061 cmÀ1; HRMS (ESI): m/z calcd for C18H32O4Na: 335.2198
[M+Na]+·; found: 335.2203.
of 2 could be linearly transferred into flow mode, thus estab-
lishing procedures that potentially allow for the production
of multi-gram quantities of chemical intermediates by con-
tinuous processing, and also at real convenience for the ex-
perimentalist. The synthetic intermediates produced in flow
showed comparable purity to those synthesised in the con-
ventional way. The improved seven-step sequence towards
isoxazolidine 18 involved only two chromatographic separa-
tions, and most products could be isolated in sufficiently
pure condition by simple filtration through a minimal
amount of silica or alumina adsorbants; aqueous workup
was unnecessary in all cases, except the aqueous hydrolysis
of acetal 14 to aldehyde 15. Moreover, we demonstrated
that flow processing, when involving the use of strong or-
ganometallic bases such as n-butyllithium, LDA and
KHMDS can be performed with equal efficiency as batch
processing, but at temperatures in the vicinity of 08C or
higher, rather than “typical” À788C.[20] The example of a Pe-
terson olefination performed at 08C in flow, which proceeds
with comparable stereoselectivity to the batch reaction in a
dry-ice bath, is a powerful demonstration of the advantage
of flow processing.
(9S)-1,1-Diethoxy-9-hydroxy-tetradecan-5-one (11): A mixture of the al-
kynone 10 (21.1 g, 67.6 mmol), Pd(OH)2 on carbon (1.3 g of 20% Pd cat-
alyst) and Et3N (2.0 mL) in EtOAc (150 mL) was shaken vigorously
under an atmosphere of H2 (25 psi, 3.5 h). The mixture was filtered
through Celite washing with EtOAc. The combined filtrate and washings
were concentrated and the residue was purified by silica gel funnel chro-
matography (EtOAc/hexanes containing 1% Et3N, gradient; 10:90, 15:85,
30:70 then 35:65) to yield alcohol 11 (20.7 g, 94%) as a colourless, waxy
solid: Rf 0.20 (25:75 EtOAc/toluene); [a]2D6 =À4.5 (c 1.06, CHCl3);
1H NMR (CDCl3, 400 MHz) d 0.87 (t, J=6.8 Hz, 3H), 1.18, (t, J=7.1 Hz,
6H), 1.22–1.82 (m, 16H), 2.42 (t, J=7.0 Hz, 4H), 3.42–3.67 (m, 4H),
4.46, (t, 5.2 Hz, 1H) ppm; 13C NMR (CDCl3, 100 MHz) d 14.0, 15.3, 19.1,
19.7, 22.6, 25.3, 31.8, 33.0, 36.8, 37.4, 42.3, 42.5, 61.1, 71.4, 102.7,
211.0 ppm; IR (neat) 3291, 3203, 2922, 2869, 1701, 1456 cmÀ1; HRMS
(ESI) m/z 339.2526 C18H36O4Na [M+Na]+· requires 339.2511.
Experimental Section
General methods: Batch reactions were performed under exclusion of air
and moisture, in flame-dried glassware and under an atmosphere of nitro-
gen or argon. Flow reactions were performed with microreactor systems
as specified in the respective experimental procedures. Microwave reac-
tions were carried out in sealed reaction vessels using a Biotage Initiator
2.0 instrument (400 W). Commercial reagents were used without further
purification, and enantiopure 8 was generously provided by ZEON Cor-
poration.[10] Anhydrous solvents were obtained by passing them through
columns of activated alumina (tetrahydrofuran, dichloromethane, tolu-
ene) or purchased from commercial suppliers (tert-butyl methyl ether,
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
32.1 mmol) and Et3N (17.8 mL, 128 mmol) in CH2Cl2 (250 mL) at À408C
was added MsCl (4.96 mL, 64.2 mmol) in CH2Cl2 (50 mL) dropwise
11476
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11471 – 11480

Total Synthesis of Perhydrohistrionicotoxin
FULL PAPER
(5 min). This mixture was allowed to warm to 08C over 1 h. Saturated
aqueous NaHCO3 was added and the organic layer was separated. The
aqueous layer was extracted with CH2Cl2 and the combined organic
layers were washed with saturated aqueous NaCl, dried (MgSO4), filtered
and concentrated. The residue was purified by silica gel funnel chroma-
tography (EtOAc/hexanes containing 1% Et3N, gradient; 5:95, 10:90,
15:85 then 25:75) to give the mesylate 12 (11.1 g) as a clear, colourless oil
which was used immediately in the next step. Analytical data: Rf 0.28
concentrated. The residue was purified by flash chromatography
(EtOAc/hexanes, gradient; 3:97, 4:96, 5:95 then 6:94) to yield a 95:5
Z/E-mixture of a,b-unsaturated nitrile 17 (4.45 g, 86%) as a clear, colour-
less oil: Rf 0.27 (10:90 EtOAc/hexanes); [a]2D4 =À13.8 (c 0.95, CHCl3);
1H NMR (CDCl3, 400 MHz) for major Z isomer d 0.88 (t, J=6.8 Hz,
3H), 1.06–1.71 (m, 15H), 1.75–1.87 (m, 2H), 1.88–2.01 (m, 2H), 2.27–
2.38 (m, 2H), 2.64–2.77 (m, 2H), 5.20 (d, J=10.9 Hz, 1H), 5.41 (dd, J=
5.0, 10.1 Hz, 1H), 6.31 (dt, J=7.6, 10.8 Hz, 1H), 7.21–7.41 (m, 5H) ppm;
13C NMR (CDCl3, 100 MHz) for major (Z-) isomer d 14.0, 19.9, 22.6,
25.3, 29.5, 31.0, 31.9, 32.1, 34.5, 41.3, 42.1, 59.4, 67.6, 76.7, 99.4, 116.0,
125.9, 127.0, 128.3, 142.0, 155.0 ppm; IR (neat) 2929, 2860, 2219, 1453,
753, 698 cmÀ1; HRMS (EI) m/z 366.2660. C24H34N2O [M]+· requires
366.2666. Characteristic signals for minor E isomer: 1H NMR (CDCl3,
400 MHz) d 5.24 (dt, J=1.6, 16.4 Hz, 1H); 6.58 (td, J=6.9, 16.4 Hz, 1H),
ppm.
(30:70 EtOAc/hexanes); 1H NMR (CDCl3, 400 MHz)
d 0.87 (t, J=
6.8 Hz, 3H), 1.18, (t, J=7.2 Hz, 6H), 1.22–1.44 (m, 6H), 1.53–1.73 (m,
10H) 2.39–2.46 (m, 4H), 2.99 (s, 3H), 3.47 (qd, J=7.1, 9.5 Hz, 2H), 3.62
(qd, J=7.1, 9.5 Hz, 2H), 4.46, (t, J=5.3 Hz, 1H), 4.69 (mc, 1H) ppm;
13C NMR (CDCl3, 100 MHz) d 13.8, 15.2, 18.8, 18.9, 22.3, 24.5, 31.4, 32.9,
33.7, 34.2, 38.6, 41.8, 42.3, 61.0, 83.4, 102.6, 210.0 ppm. Mesylate 12
(11.1 g) was taken up in EtOH (250 mL) and NaHCO3 (14.7 g,
174 mmol) and NH2OH·HCl (11.7 g, 168 mmol) were added sequentially.
This mixture was stirred at RT (30 min), then at 708C (20 h). The reac-
tion mixture was allowed to cool to RT, H2O (200 mL) was added and
the mixture was concentrated to approximately 200 mL volume. The mix-
ture was diluted with CH2Cl2 and the organic layer was separated. The
aqueous layer was extracted (4 ꢃ) with CH2Cl2 and the combined organ-
ic layers were washed with saturated aqueous NaCl, dried (MgSO4), fil-
tered and concentrated to afford crude nitrone 13 (10.1 g) as a pale
yellow oil which was used immediately in the next step without character-
isation, Rf 0.34 (4:96 EtOH/CH2Cl2). Crude nitrone 13 (10.1 g) was taken
up in freshly distilled styrene (200 mL) and heated at 708C (72 h). The
mixture was allowed to cool to RT, concentrated and the residue was pu-
rified by flash chromatography (EtOAc/hexanes containing 1% Et3N,
gradient; 2:98, 4:96 then 6:94) to yield the isoxazolidine exo-14 (5.89 g,
44%) as a colourless oil: Rf 0.33 (10:90 EtOAc/hexanes); [a]2D5 =À16.2 (c
(1R,5R,8S,12R)-12-Cyano-5-pentyl-6-aza-7-oxatricyclo[6.3.1.01,6]dode-
cane (À)-18: A solution of the (95:5, Z/E) a,b-unsaturated nitrile 17
(4.03 g, 11.0 mmol) in PhCH3 (114 mL) and EtOH (19 mL) was divided
equally into 19 vials. These vials were each sealed and irradiated in a mi-
crowave reactor for 50 min at 1848C (absorption level “normal”, the
pressure was 11 bar). The contents of the vials were combined, concen-
trated and the residue was purified by flash chromatography (EtOAc/
hexanes, gradient; 3:97, 4:96 then 5:95) to yield the tricycle 18 (2.03 g,
70%) as a colourless oil: Rf 0.23 (8:92 EtOAc/hexanes) [a]2D4 =À201.9 (c
1.10, CHCl3); 1H NMR (CDCl3, 400 MHz) d 0.86 (t, J=6.8 Hz, 3H),
1.06–1.40 (m, 9H), 1.50–1.70 (m, 4H), 1.71–2.04 (m, 6H), 2.11–2.20 (m,
1H), 2.33–2.43 (m, 1H), 3.43 (d, J=5.5 Hz, 1H), 4.64–4.75 (m, 1H) ppm;
13C NMR (CDCl3, 100 MHz) d 14.0, 17.5, 19.2, 22.6, 25.3, 27.1, 29.6, 32.0,
32.2, 34.1, 36.0, 38.1, 65.5, 65.5, 75.6, 117.9 ppm; IR (KBr disc) 2939,
2861, 2236, 1456, 1082, 930 cmÀ1; HRMS (EI) m/z 262.2035. C16H26N2O
[M]+· requires 262.2040.
1
1.25, CHCl3); H NMR (CDCl3, 400 MHz) d 0.87 (t, J=6.8 Hz, 3H), 1.14
(t, J=6.8 Hz, 3H), 1.15 (t, J=6.8 Hz, 3H), 1.17–1.70 (m, 16H), 1.75–1.97
(m, 3H), 2.00 (dd, J=5.2, 12.5 Hz, 1H), 2.63, (t, J=11.1 Hz, 1H), 2.67–
2.77 (m, 1H), 3.31–3.47 (m, 2H), 3.49–3.61 (m, 2H), 4.39, (t, J=5.6 Hz,
1H), 5.39, (dd, J=5.0, 10.0 Hz, 1H), 7.18–7.41 (m, 5H) ppm; 13C NMR
(CDCl3, 100 MHz) d 14.0, 15.3, 18.9, 20.0, 22.6, 25.4, 29.5, 31.3, 32.1, 33.9,
34.6, 41.5, 42.2, 59.3, 60.7, 61.0, 67.8, 76.9, 102.6, 126.0, 127.0, 128.3,
Flow- and microwave-assisted synthesis of 6,6,5-tricyclic isoxazolidine (Æ
)-18
Telescoped flow preparation of propargyl ketone 10: A solution of
alkyne 9 (4.26 g, 30.0 mmol) in THF was diluted to 100 mL in a volumet-
ric flask (resulting in a 0.30m solution). Likewise, a solution of nBuLi
(16.3 mL, 1.60m in hexane, 26.1 mmol) in hexane was diluted to 100 mL
in a volumetric flask (resulting in a 0.26m solution), and a solution of lac-
tone (Æ)-8 (0.43 g, 2.53 mmol) in THF, diluted to 25 mL in a volumetric
flask (resulting in a 0.10m solution). The stock solutions of nBuLi and
142.0 ppm; IR (neat) 2929, 2865, 1450, 1374, 1125, 1063, 754, 698 cmÀ1
;
C38H53NO4Si requires C, 74.10; H, 8.67; N, 2.27, found: C, 74.17; H, 8.63;
N, 2.29; HRMS (EI) m/z 417.3220. C26H43NO3 [M]+· requires 417.3237.
ACHTUNGTRENNUNG
alkyne
9 ) at a rate of
were pumped (piston pumps, Ismatec[13]
ACHTUNGTRENNUNG
450 mLminÀ1 each, onto a glass chip (internal volume 0.8 mL, Sigma Al-
drich[15]), where mixing occurred at 08C. The resulting flow stream was
directed into a 30 mL reactor coil (PTFE tubing, 1/8 in. o.d., 1.5 mm i.
d.), cooled to 08C, then onto a second identical glass chip. As soon as a
constant flow stream of the lithiated alkyne entered the second glass
chip, evident by residence time calculation and the light yellow colour of
the solution, it was combined at 08C with a flow stream of lactone (Æ)-8
(pumped at 900 mLminÀ1 with a Knauer HPLC pump[14]). The resulting
flow stream was passed through a second reactor coil (10 mL), cooled to
08C and into an aqueous solution of NH4Cl at the reactor outlet. After
the stock solution of the lactone had been consumed, rinse was per-
formed using THF until all reagent had exited the reactor (residual stock
solutions of alkyne and nBuLi were left unreacted). The crude product
mixture was transferred into a separation funnel, the layers were separat-
ed and the aqueous layer was extracted with EtOAc (3 ꢃ). The com-
bined organic layers were dried (MgSO4), filtered, evaporated and drying
in vacuum afforded ketone 10 (0.65 g, 83%) as a light yellow oil, purity
85–90% as estimated by 1H NMR spectroscopy, the product being spec-
troscopically identical to the product obtained by the batch procedure.
14.1 mmol) in THF (100 mL) was added aqueous HCl (20.0 mL of a 2.0m
solution, 40.0 mmol). After 1.5 h, saturated aqueous NaHCO3 (40 mL)
was added. The mixture was further diluted with H2O and the organic
layer was separated. The aqueous layer was extracted with EtOAc (2 ꢃ)
and the combined organic layers were washed with saturated aqueous
NaCl, dried (MgSO4), filtered and concentrated. The residue was filtered
through a plug of silica, washing with EtOAc/hexanes (25:75). The fil-
trate and washings were concentrated to give the aldehyde 15 (5.55 g) as
a colourless oil which was used immediately in the next step. Analytical
1
data: Rf 0.41 (20:80 EtOAc/hexanes); H NMR (CDCl3, 400 MHz) d 0.81
(t, J=6.8 Hz, 3H), 1.07 (mc, 1H), 1.14–1.80 (m, 16H), 1.85 (mc, 1H), 1.90
(dd, J=5.2, 12.5 Hz, 1H), 2.23 (dddd, J=1.7, 6.3, 8.1, 17.4 Hz, 1H), 2.34
(dddd, J=1.7, 6.3, 8.1, 17.4 Hz, 1H), 2.62 (dd, J=10.3, 12.5 Hz, 1H), 2.66
(tt, J=3.2, 8.2 Hz, 1H), 5.33 (dd, J=5.2, 10.3 Hz, 1H), 7.19–7.14 (m,
1H), 7.23–7.33 (m, 4H), 9.60 (t, J=1.7 Hz, 1H) ppm; 13C NMR (CDCl3,
100 MHz) d 14.0, 16.6, 19.9, 22.7, 25.4, 29.4, 31.0, 32.2, 34.5, 41.6, 42.2,
44.2, 59.3, 67.7, 76.8, 125.9, 127.1, 128.3, 141.9, 202.7 ppm. Meanwhile, to
a solution of (tBuO)Ph2SiCH2CN[12] (5.20 g, 17.6 mmol) in THF (150 mL)
at À788C was added nBuLi (11.0 mL, 1.60m in hexanes, 17.6 mmol)
dropwise (over ꢀ1 min). The mixture was allowed to warm to 08C, kept
at this temperature for 10 min, and then cooled to À788C. To this mix-
ture was added dropwise (over ꢀ10 min) the crude aldehyde (5.55 g) in
THF (50 mL). After 1 h, saturated aqueous NH4Cl (40 mL) was added
and the mixture was allowed to warm to RT. The mixture was diluted
with H2O and EtOAc and the organic layer was separated. The aqueous
layer was extracted with EtOAc (2 ꢃ) and the combined organic layers
were washed with saturated aqueous NaCl, dried (MgSO4), filtered and
Flow hydrogenation of propargyl ketone 10 to alcohol 11: Propargyl
1
ketone 10 (386 mg containing ꢀ10% of alkyne 9 by H NMR spectrosco-
py, ꢀ1.17 mmol) was dissolved in EtOAc (45 mL) and iPr2NEt (0.90 mL)
and the solution was passed through the H-cube hydrogenator [Thales
Nano,[18] catalyst 20% Pd(OH)2/C, mode “full H2”] at 1 mLminÀ1, the
catalyst being heated at 408C. The product mixture was collected at the
reactor outlet, evaporated and vacuum-dried to afford pure alcohol 11
Chem. Eur. J. 2010, 16, 11471 – 11480
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11477

J. H. Ryan et al.
(354 mg, 96%) as a colourless oil, spectroscopically identical to the prod-
uct obtained by the batch procedure.
tions of 2 mL volume each were injected into the sample loops connected
to three HPLC pumps. A solution of KHMDS (0.50m in PhCH3, 2 mL)
was pumped at a rate of 120 mLminÀ1 into a Y-shaped mixer piece
(PEEK), where it was combined with a solution of silyl reagent 19
Flow mesylation of alcohol 11 to mesylate 12: MsCl (0.12 mL,
1.55 mmol) was diluted to a 10 mL volume with CH2Cl2 in a volumetric
flask. Likewise, a CH2Cl2 solution of alcohol 11 (316 mg, 1.00 mmol) and
iPr2NEt (0.35 mL, 2.00 mmol) was diluted to a 10 mL volume in a second
volumetric flask. The two stock solutions were pumped (piston pumps, Is-
matec[13]) at a rate of 450 mLminÀ1 each, onto a glass chip (internal
volume 0.8 mL, Sigma Aldrich[15]), where mixing occurred at RT. After
the stock solutions had been consumed, rinse was performed with
CH2Cl2 until all reagents had exited the reactor. The product mixture
was collected at the reactor outlet and evaporated onto basic alumina.
The crude product was charged onto a short plug of basic alumina and
flushed with Et2O. The filtrate was concentrated and dried in vacuum to
afford pure mesylate 12 (390 mg, 99%) as a colourless oil, spectroscopi-
cally identical to the product obtained by the batch procedure.
(305 mg, 0.78 mmol in 2 mL PhCH3, 0.39m), pumped at
a rate of
200 mLminÀ1. The resulting flow stream was directed into a 5 mL reactor
coil, cooled to 08C, and into a second Y piece thereafter, connected to a
second 5 mL reactor coil. After the calculated residence time of the de-
protonated reagent 19 in the first reactor coil had passed, a solution of al-
dehyde 15 (68.0 mg, 0.20 mmol in 2 mL PhCH3, 0.10m) was channelled in
through the second Y piece, and the resulting flow stream passed through
the second 5 mL reactor coil at 08C. As the aldehyde solution was con-
sumed (by time calculation), the first two reaction channels were
switched off, and the third pump was set to 320 mLminÀ1, and rinse was
performed until all material had exited the reactor through the omnifit
cartridge. The crude product was collected, evaporated onto silica and
charged onto
a short plug of silica. This was flushed with CH2Cl2
One-pot microwave reaction of mesylate 12 to isoxazolidine 14: A mix-
ture of mesylate 12 (395 mg, 1.00 mmol), iPr2NEt (2.10 mL, 12.00 mmol)
and H2NOH·HCl (556 mg, 8.00 mmol) in styrene (11 mL) was stirred in a
20 mL microwave vial until homogeneous (ꢀ15 min), then heated in the
microwave at 1058C for 6 h (absorption level “high”). The mixture was
filtered through a plug of basic alumina with the aid of EtOAc, concen-
trated and vacuum-dried. The residue was purified by chromatography
(EtOAc/hexanes 1:8 containing 1% iPr2NEt) to afford isoxazolidine 14
(166 mg, 40%) as a 85:15 mixture of exo and endo isomers, the major
(separable) component being spectroscopically identical to the product
obtained by the batch procedure described above. The minor component,
the endo isomer,[21] could not be obtained in pure form. Its characteristic
NMR signals were obtained from a mixture fraction: endo-14: Rf 0.40
(10:90 EtOAc/hexanes); characteristic signals: 1H NMR (CDCl3,
400 MHz) d 4.53, (t, J=5.6 Hz, 1H), 5.23, (t, J=8.7 Hz, 1H) ppm;
13C NMR (CDCl3, 100 MHz) d 14.1, 15.4, 19.3, 19.5, 22.7, 25.0, 29.68,
29.74, 30.9, 34.2, 34.3, 40.3, 40.6, 60.9, 61.3, 62.5, 67.6, 79.9, 102.8, 125.2,
126.7, 128.1, 144.6 ppm.
(100 mL), the filtrate was discarded, then with EtOAc (100 mL). The
EtOAc flushings were evaporated and dried in vacuum to provide pure
nitrile 17 as a light yellow oil (61.0 mg, 83%), showing a Z/E ratio of
>90:10 by 1H NMR spectroscopy and was spectroscopically identical to
the product obtained by the batch procedure.
Flow rearrangement of nitrile 17 to isoxazolidine (Æ)-18: A solution of
nitrile 17 (61.0 mg, 166 mmol) in PhCH3 (2 mL) was injected into a 2 mL
sample loop (Vapourtec R2+/R4[19]), and passed through a steel reactor
coil (10 mL) at 500 mLminÀ1 heated to 1908C. The product mixture was
collected at the reactor outlet, evaporated and purified by chromatogra-
phy (EtOAc/hexanes 1:10) to afford isoxazolidine (Æ)-18 as a colourless
oil (27.0 mg, 62%), spectroscopically identical to the product obtained by
the batch procedure.
Conversion of 6,6,5-tricyclic isoxazolidine (À)-18 into 2
(1R,5R,8S,12S)-(1’Z)-12-(But-1’-enyl)-5-pentyl-6-aza-7-oxatricy-
clo[6.3.1.01,6]dodecane 20: iBu2AlH (9.14 mL, 1.50m in PhCH3,
13.7 mmol) was added to
a
solution of the nitrile (À)-18 (1.80 g,
Preparation of silyl reagent 19, 2-[(2-methylundecan-2-yloxy)diphenylsi-
lyl]acetonitrile: Ph2SiCl2 (1.79 mL, 8.51 mmol) was added at RT to a mix-
ture of 2-methylundecan-2-ol[22] (1.58 g, 8.50 mmol) and iPr2NEt
(1.53 mL, 8.78 mmol) in CH2Cl2 (25 mL) and the mixture was stirred at
reflux for 4 days. The mixture was concentrated to dryness and the resi-
due was taken up in petroleum ether (100 mL) and stirred for 15 min.
The mixture was filtered through Celite with the aid of petroleum ether
and the filtrate was concentrated and vacuum dried to afford the inter-
mediate monochlorosilane as a colourless oil (2.11 g, 62%), which was
used immediately in the next step without characterisation. nBuLi
(6.55 mL, 1.60m in hexane, 10.48 mmol) was added at 08C to a solution
of iPr2NH (1.52 mL, 10.84 mmol) in THF (12 mL) and the mixture was
stirred for 30 min. Dry MeCN (0.29 mL, 5.55 mmol) was added and the
mixture was stirred at 08C for 1 h. A solution of the chlorosilane (2.11 g,
5.23 mmol) in THF (5 mL) was added and the mixture was stirred at RT
for 2 h. After addition of an aqueous solution of NH4Cl (20 mL), the
layers were separated and the aqueous layer was extracted with Et2O
(3ꢃ). The combined organic layers were dried (MgSO4), filtered, concen-
trated and purified by chromatography (petroleum ether/CH2Cl2 1:1) to
give the silyl reagent 19 as a colourless oil (1.65 g, 80%). Rf 0.55 (50:50
petroleum ether/CH2Cl2); 1H NMR (CDCl3, 400 MHz): d=0.89 (t, J=
6.8 Hz, 3H), 1.17–1.42 (m, 20H), 1.46–1.51 (m, 2H), 2.14 (s, 2H), 7.38–
7.50 (m, 6H), 7.62–7.66 ppm (m, 4H); 13C NMR (CDCl3, 100 MHz): d=
5.7, 14.1, 22.7, 24.4, 29.3, 29.56, 29.58, 29.8, 30.0, 31.9, 44.6, 77.3, 118.4,
128.1, 130.6, 133.8, 134.6 ppm; IR (thin film) 2924, 2853, 2236, 1466, 1428,
1384, 1367, 1146, 1112, 1046, 1027 cmÀ1; HRMS (EI) m/z calcd for
C26H37NOSi: 407.2644 [M]+·; found: 407.2629.
6.85 mmol) in PhCH3 (50 mL) at À788C. After 4 h, MeOH (1 mL) was
added and the mixture was allowed to warm to RT. The mixture was di-
luted with EtOAc (50 mL) and aqueous potassium sodium tartrate (1.4m,
100 mL) and stirred vigorously (4 h). The organic layer was separated
and the aqueous layer was extracted with EtOAc (2ꢃ) and the combined
organic layers were washed with saturated aqueous NaCl, dried
(MgSO4), filtered and concentrated. The residue was filtered through a
plug of silica, washing with EtOAc/hexanes (25:75). The filtrate and
washings were concentrated to give the intermediate aldehyde (1.85 g) as
a colourless oil [Rf 0.37 (20:80 EtOAc/hexanes)] that was used immedi-
ately. Meanwhile, to a mixture of propyltriphenylphosphonium bromide
(3.96 g, 10.3 mmol) in THF (20 mL) at À788C was added nBuLi
(6.47 mL, 1.59m in hexanes, 10.3 mmol) dropwise (over ꢀ3 min). The
light yellow mixture was allowed to warm to 08C and stirred for 30 min
whereupon the mixture had turned orange. The mixture was cooled to
À788C and the crude aldehyde (1.85 g) in THF (30 mL) was added drop-
wise (over ꢀ20 min). This mixture was allowed to slowly warm to RT
(over ꢀ4 h) and stirred overnight. A saturated aqueous solution of
NH4Cl (20 mL) was added and the mixture was diluted with H2O and
EtOAc and the organic layer was separated. The aqueous layer was ex-
tracted (2ꢃ) with EtOAc and the combined organic layers were washed
sequentially with saturated aqueous solutions of NaHCO3 and NaCl,
dried (MgSO4), filtered and concentrated. The residue was purified by
flash chromatography (EtOAc/hexanes; 2:98, 4:96 then 5:95) to yield the
Z-alkene 20 as a clear, colourless oil (1.11 g, 55%). Rf 0.29 (8:92 EtOAc/
hexanes); [a]2D4 =À160.8 (c 1.06, CHCl3); 1H NMR (CDCl3, 400 MHz):
d=0.86 (t, J=6.5 Hz, 3H), 0.98 (t, J=7.5 Hz, 3H), 1.04–1.63 (m, 16H),
1.7–1.82 (m, 2H), 1.85–2.02 (m, 2H), 2.04–2.18 (m, 2H), 2.56–2.68 (m,
1H), 3.42 (dd, Jꢀ7.2 Hz, 1H), 4.32–4.39 (m, 1H), 5.42 (t, J=10.5 Hz,
1H), 5.62 ppm (dt, J=11.0, 7.3 Hz, 1H); 13C NMR (CDCl3, 100 MHz):
d=14.0, 14.4, 17.9, 19.6, 21.2, 22.6, 25.0, 25.7, 29.8, 32.2, 32.3, 34.2, 34.7,
42.5, 64.8, 64.8, 77.9, 123.3, 136.2 ppm; IR (thin film): 3022, 2928, 2858,
Telescoped Peterson olefination of aldehyde 15 with silyl reagent 19 to
nitrile 17: Two Vapourtec R2+/R4 units[19] were connected as shown in
Scheme 5 (also see Figure S2 in the Supporting Information) and two
5 mL reactor coils (PTFE tubing, 1/16 in. o.d., 1 mm i. d.), separated by a
one-way check valve, were appended to the system, as well as an omnifit
column (150 mm, 10 mm bore)[16] filled with QP-IDA (5.50 g,
0.16 mmolgÀ1). The system was primed with PhCH3. Three reagent solu-
1459, 926 cmÀ1; HRMS (EI): m/z calcd for C19H33NO: 291.2557 [M]+·
;
found: 291.2557.
11478
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11471 – 11480

Total Synthesis of Perhydrohistrionicotoxin
FULL PAPER
(À)-Perhydrohistrionicotoxin 2: A mixture of the alkene 20 (1.05 g,
3.61 mmol), aqueous HCl (8.00 mL, 1.00m, 8.00 mmol) and Pd(OH)2 on
carbon (200 mg of 20% Pd catalyst) in THF (150 mL) was shaken vigo-
rously under an atmosphere of H2 (25 psi, 14 h). The mixture was filtered
through Celite, washed with THF and then THF/H2O (1:1). The com-
bined filtrate and washings were concentrated and the residue was dilut-
ed with a saturated aqueous solution of NaHCO3 and EtOAc. The organ-
ic layer was separated and the aqueous layer was extracted with EtOAc
(2ꢃ). The combined organic layers were washed with a saturated aque-
ous solution of NaCl, dried (MgSO4), filtered and concentrated. The resi-
due was purified by flash chromatography (EtOAc/hexanes; 30:70 then
MeOH/CH2Cl2/aqueous NH3 (14m); 2:98:0, 2:97:1, 5:94:1 then 10:89:1)
to afford 2 as a waxy colourless solid (953 mg, 89%). Rf 0.28 (10:90
MeOH/CH2Cl2); [a]D21 =À64 (c 0.97, CHCl3; lit. [24] À84.1);[23] 1H NMR
(CDCl3, 500 MHz): d=0.76 (td, J=13.8, 3.8 Hz, 1H), 0.86 (t, J=7.2 Hz,
3H), 0.90 (t, J=6.8 Hz, 3H), 1.04–1.17 (m, 2H), 1.20–1.43 (m, 15H),
1.44–1.66 (m, 5H), 1.73 (m, 1H), 1.81 (dt, J=13.8, 2.5 Hz, 1H), 2.03 (qt,
J=13.8, 4.2 Hz, 1H), 2.20 (d, J=9.7 Hz, 1H), 2.92 (m, 1H), 3.92 ppm (d,
J=2.5 Hz, 1H); 13C NMR (CDCl3, 100 MHz): d=14.0, 14.1, 15.2, 19.5,
22.5, 23.0, 25.6, 27.6, 27.7, 30.3, 32.1, 33.1, 36.8, 36.9, 37.7, 38.1, 50.0, 55.3,
70.0 ppm; all data are in close agreement with the literature;[24] IR (thin
film) 3254, 2928, 2859, 1457, 1135, 1067, 976 cmÀ1; HRMS (EI): m/z calcd
for C19H37NO: 295.2870 [M]+·; found: 295.2871. A small portion was
treated with an excess of HCl in MeOH (0.1m) and then concentrated to
give (À)-perhydrohistrionicotoxin (HCl salt) 2·HCl [a]2D1 =À34.5 (c 1.06,
CHCl3; lit. [4] À34.5).
[2] For selected recent examples of the development of flow methodol-
ogy, see: a) M. Baumann, I. R. Baxendale, L. J. Martin, S. V. Ley,
Tetrahedron 2009, 65, 6611–6625; b) I. R. Baxendale, S. V. Ley, A.
Mansfield, C. D. Smith, Angew. Chem. 2009, 121, 4077–4081;
Angew. Chem. Int. Ed. 2009, 48, 4017–4021; c) K. Mennecke, A.
Kirschning, Beilstein J. Org. Chem. 2009, 5, No. 21; d) B. Ahmed-
Omer, D. A. Barrow, T. Wirth, Tetrahedron Lett. 2009, 50, 3352–
3355; e) A. Odedra, P. H. Seeberger, Angew. Chem. 2009, 121,
2737–2740; Angew. Chem. Int. Ed. 2009, 48, 2699–2702; f) J. C.
Brandt, T. Wirth, Beilstein J. Org. Chem. 2009, 5, No. 30; g) R. Kik-
keri, P. Laurino, A. Odedra, P. H. Seeberger, Angew. Chem. 2010,
122, 2098–2101; Angew. Chem. Int. Ed. 2010, 49, 2054–2057;
h) C. F. Carter, I. R. Baxendale, J. B. J. Pavey, S. V. Ley, Org.
Biomol. Chem. 2010, 8, 1588–1595; i) L. Tamborini, P. Conti, A.
Pinto, C. De Micheli, Tetrahedron: Asymmetry 2010, 21, 222–225.
[3] For examples of telescoped multi-step flow synthesis, see: a) I. R.
Baxendale, J. Deeley, C. M. Griffiths-Jones, S. V. Ley, S. Saaby, G. K.
Tremner, Chem. Commun. 2006, 2566–2568; b) I. R. Baxendale,
C. M. Griffiths-Jones, S. V. Ley, G. K. Tranmer, Synlett 2006, 427–
430; c) T. Gustafsson, F. Pontꢆn, P. H. Seeberger, Chem. Commun.
2008, 1100–1102; d) T. P. Petersen, A. Ritzꢆn, T. Ulven, Org. Lett.
2009, 11, 5134–5137; e) Z. Qian, S. V. Ley, Synlett 2010, 505–508;
f) I. R. Baxendale, S. C. Schou, J. Sedelmeier, S. V. Ley, Chem. Eur.
J. 2010, 16, 89–94.
[4] Isolation: J. W. Daly, I. Karle, C. W. Myers, T. Tokuyama, J. A.
Waters, B. Witkop, Proc. Natl. Acad. Sci. USA 1971, 68, 1870–1875.
[5] For a review, see: A. Sinclair, R. A. Stockman, Nat. Prod. Rep. 2007,
24, 298–326, and references therein.
[6] a) G. M. Williams, S. D. Roughley, J. E. Davies, A. B. Holmes, J. P.
Adams, J. Am. Chem. Soc. 1999, 121, 4900–4901; b) C. J. Smith,
A. B. Holmes, N. J. Press, Chem. Commun. 2002, 1214–1215;
c) E. C. Davison, M. E. Fox, A. B. Holmes, S. D. Roughley, C. J.
Smith, G. M. Williams, J. E. Davies, P. R. Raithby, J. P. Adams, I. T.
Forbes, N. J. Press, M. J. Thompson, J. Chem. Soc. Perkin Trans. 1
2002, 1494–1514; d) H. T. Horsley, A. B. Holmes, J. E. Davies, J. M.
Goodman, M. A. Silva, S. I. Pascu, I. Collins, Org. Biomol. Chem.
2004, 2, 1258–1265; e) J. M. Macdonald, H. T. Horsley, J. H. Ryan,
S. Saubern, A. B. Holmes, Org. Lett. 2008, 10, 4227–4229.
[7] a) A. J. Lapa, E. X. Albuquerque, J. Sarvey, J. W. Daly, B. Witkop,
Exp. Neurol. 1975, 47, 558–580; b) C. E. Spivak, M. A. Maleque,
A. C. Oliveria, L. M. Masukawa, T. Tokuyama, J. W. Daly, E. X. Al-
buquerque, Mol. Pharmacol. 1982, 21, 351–361; c) J. W. Daly, Y.
Nishizawa, J. A. Edwards, J. A. Waters, R. S. Aronstam, Neurochem.
Res. 1991, 16, 489–500.
[8] a) R. A. Stockman, Tetrahedron Lett. 2000, 41, 9163–9165; b) L. G.
Arini, P. Szeto, D. L. Hughes, R. A. Stockman, Tetrahedron Lett.
2004, 45, 8371–8374; c) R. A. Stockman, A. Sinclair, L. G. Arini, P.
Szeto, D. L. Hughes, J. Org. Chem. 2004, 69, 1598–1602; d) M. S.
Karatholuvhu, A. Sinclair, A. F. Newton, M. L. Alcaraz, R. A.
Stockman, P. L. Fuchs, J. Am. Chem. Soc. 2006, 128, 12656–12657.
[9] M. Brasholz, B. A. Johnson, J. M. Macdonald, A. Polyzos, J. Tsa-
naktsidis, S. Saubern, A. B. Holmes, J. H. Ryan, Tetrahedron, 2010,
66, 6445–6449.
Acknowledgements
The authors thank the Commonwealth Scientific and Industrial Research
Organisation (Office of the Chief Executive) for a postdoctoral fellow-
ship (MB) and the Australian Research Council for financial support
(DP 0451189).
[1] For books and book chapters on the use of flow reactors in synthe-
sis, see: a) New Avenues to Efficient Chemical Synthesis: Emerging
Technologies. Ernst Schering Foundation Symposium Proceedings,
Vol. 1 (Eds.: P. H. Seeberger, T. Blume), Springer, Berlin, Heidel-
berg, 2007; b) Microreactors in Organic Synthesis and Catalysis (Ed.:
T. Wirth), Wiley-VCH, Weinheim, 2008; c) J. C. Brandt, T. Wirth in
Recoverable and Recyclable Catalysts (Ed.: M. Benaglia), Wiley,
2009, 411–425; d) S. Ceylan, A. Kirschning in Recoverable and Re-
cyclable Catalysts (Ed.: M. Benaglia), Wiley, 2009, 379–410;
e) Chemical Reactions and Processes under Flow Conditions, RSC
Green Chemistry Series No. 5 (Eds.: S. V. Luis, E. Garcia-Verdugo),
The Royal Society of Chemistry, Cambridge, 2010; for review arti-
cles, see: f) G. Jas, A. Kirschning, Chem. Eur. J. 2003, 9, 5708–5723;
g) K. Jꢄhnisch, V. Hessel, H. Lçwe, M. Baerns, Angew. Chem. 2004,
116, 410–451; Angew. Chem. Int. Ed. 2004, 43, 406–446; h) A.
Kirschning, W. Solodenko, K. Mennecke, Chem. Eur. J. 2006, 12,
5972–5990; i) B. Ahmed-Omer, J. C. Brandt, T. Wirth, Org. Biomol.
Chem. 2007, 5, 733–740; j) B. P. Mason, K. E. Price, J. L. Steinbach-
er, A. R. Bogdan, D. T. McQuade, Chem. Rev. 2007, 107, 2300–
2318; k) V. T. N. Glasnov, C. O. Kappe, Macromol. Rapid Commun.
2007, 28, 395–410; l) F. Benito-Lꢅpez, R. J. M. Egberink, D. N.
Reinhoudt, W. Verboom, Tetrahedron 2008, 64, 10023–10040;
m) S. V. Ley, I. R. Baxendale, Chimia 2008, 62, 162–168; n) L. F.
Raveglia, G. A. M. Giardina, Future Med. Chem. 2009, 1, 1019–
1023; o) C. Wiles, P. Watts, Future Med. Chem. 2009, 1, 1593–1612;
p) X. Y. Mak, P. Laurino, P. H. Seeberger, Beilstein J. Org. Chem.
2009, 5, No. 19; q) P. H. Seeberger, Nat. Chem. Biol. 2009, 5, 368–
372; r) F. E. Valera, M. Quaranta, A. Moran, J. Blacker, A. Arm-
strong, J. T. Cabral, D. G. Blackmond, Angew. Chem. 2010, 122,
2530–2537; Angew. Chem. Int. Ed. 2010, 49, 2478–2485.
[10] (S)-(À)-6-Pentyltetrahydropyran-2-one 8 was generously provided
by ZEON Corporation, www.zeon.co.jp, and was prepared with
97.7% ee by the CPF method, see: H. Nemoto, Tetrahedron Lett.
1994, 35, 7785–7788.
[11] N. Hꢆnaff, A. Whiting, Tetrahedron 2000, 56, 5193–5204.
[12] S. Kojima, T. Fukuzaki, A. Yamakawa, Y. Murai, Org. Lett. 2004, 6,
3917–3920.
[13] Model “REGLO-CPF digital”, www.ismatec.com.
[14] Model “Smartline 100”, www.knauer.net.
[15] Component of the “Microreactor Explorer Kit”, www.sigmaaldrich.
com.
[16] For information on Omnifit products, see www.omnifit.com.
[17] For information on QuadraPure products, see www.reaxa.com.
[18] H-cube Continuous Flow Hydrogenation Reactor”, www.thalesnano.
com.
Chem. Eur. J. 2010, 16, 11471 – 11480
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
11479

J. H. Ryan et al.
[19] For detail on the “R2+/R4” flow synthesis platform, see www.
vapourtec.co.uk.
[21] For a related example of a 1,3-dipolar cycloaddition, see: J. Markan-
du, H. I. Dondas, M. Frederickson, R. Grigg, Tetrahedron 1997, 53,
13165–13176.
[22] J. Beger, M. Meerbote, J. Prakt. Chem. 1985, 327, 2–9.
[23] The optical rotation [a]2D1 of 2 prepared in this study differs from the
value given in ref. [24], but matches well to the one obtained by us
in ref. [6e], in which we presented strong evidence to support that
this is an accurate optical rotation. It is notable that the optical rota-
tion of the HCl salt of 2 matches well with the value given in
ref. [4].
[20] The use of n-butyllithium under flow conditions in a microreactor
has been reported previously, however, mostly in lithium–halogen
exchange reactions with short retention times of the reagent; for ex-
amples, see: a) T. Schwalbe, V. Autze, M. Hohmann, W. Stirner,
Org. Process Res. Dev. 2004, 8, 440–454; b) H. Usutani, Y. Tomida,
A. Nagaki, H. Okamoto, T. Nokami, J.-I. Yoshida, J. Am. Chem.
Soc. 2007, 129, 3046–3047; c) A. Nagaki, H. Kim, J.-I. Yoshida,
Angew. Chem. 2008, 120, 7951–7954; Angew. Chem. Int. Ed. 2008,
47, 7833–7836; d) A. Nagaki, N. Takabayashi, Y. Tomida , J.-I. Yosh-
ida, Org. Lett. 2008, 10, 3937–3940; e) Y. Tomida, A. Nagaki, J.-I.
Yoshida, Org. Lett. 2009, 11, 3614–3617.
[24] J. D. Winkler, P. M. Hershberger, J. Am. Chem. Soc. 1989, 111,
4852–4856.
Received: May 24, 2010
Published online: September 8, 2010
11480
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 11471 – 11480