Synthesis of sparteine-like chiral diamines and evaluation in the enantioselective lithiation-substitution of N-(tert-butoxycarbonyl)-pyrrolidine

Article abstract of DOI:10.1039/b308410h

Three chiral diamines were synthesised and evaluated as sparteine surrogates in the lithiation-substitution of N-(tert-butoxycarbonyl)pyrrolidine. The synthesis and attempted resolution of sparteine-like diamines {(1S*, 2R*, 8R*)-10-methyl-6,10-diazatricyclo [6.3.1.0^2,6]dodecane and (1S*, 2R*, 9R*)-11-methyl-7,11-diazatricyclo[7.3.1.0^2,7]tridecane} (via inclusion complex formation) are reported. Unfortunately, it was only possible to resolve the diazatricyclo-[7.3.1.0^2,7] tridecane compound. An alternative route to (1R,2S,9S)-11-methyl-7,11-diazatricyclo[7.3.1.0^2.7]tridecane starting from the natural product, (-)-cytisine, is described. This simple three-step route furnished gram-quantities of a (+)-sparteine surrogate. X-ray crystallography of an intermediate in the route, (1R,5S,12S)-3-methoxycarbonyl-decahydro-1,5-methanopyrido [1,2-a][1,5]diazocin-8-one, enabled the stereochemistry of all of the tricyclic diamines described in this paper to be unequivocally established. Two other diamines, starting from (S)-proline and resolved 2-piperidine ethanol, were prepared using standard methods. These diamines lacked the bispidine framework of (-)-sparteine and were found to impart vastly inferior enantioselectivity. It was concluded that, for the asymmetric lithiation-substitution of N-Boc pyrrolidine, a rigid bispidine framework and only three of the four rings of (-)-sparteine are needed for high enantioselectivity. Furthermore, it is shown that diamine (1R,2S,9S)-11-methyl-7,11-diazatricyclo [7.3.1.0^2,7]tridecane is the first successful (+)-sparteine surrogate.

Full text of DOI:10.1039/b308410h

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis of sparteine-like chiral diamines and evaluation in the
enantioselective lithiation–substitution of N-(tert-butoxycarbonyl)-
pyrrolidine†
Jean-Paul R. Hermet,a David W. Porter,a Michael J. Dearden,a Justin R. Harrison,a
Tobias Koplin,a Peter O’Brien,*a Jérôme Parmene,a Vladimir Tyurin,a Adrian C. Whitwood,‡a
John Gildayb and Neil M. Smithc
a Department of Chemistry, University of York, Heslington, York, UK YO10 5DD.
E-mail: paob1@york.ac.uk; Fax: ϩ44 1904 432165; Tel: ϩ44 1904 432535
b AstraZeneca, Process R & D, Avlon Works, Severn Road, Hallen, Bristol, UK BS10 7ZE
c GlaxoSmithKline, Medicines Research Centre, Gunnels Wood Road, Stevenage, UK SG1 2NY
Received 22nd July 2003, Accepted 18th September 2003
First published as an Advance Article on the web 7th October 2003
Three chiral diamines were synthesised and evaluated as sparteine surrogates in the lithiation–substitution of N-(tert-
butoxycarbonyl)pyrrolidine. The synthesis and attempted resolution of sparteine-like diamines {(1S*,2R*,8R*)-10-
methyl-6,10-diazatricyclo[6.3.1.02,6]dodecane and (1S*,2R*,9R*)-11-methyl-7,11-diazatricyclo[7.3.1.02,7]tridecane}
(via inclusion complex formation) are reported. Unfortunately, it was only possible to resolve the diazatricyclo-
[7.3.1.02,7]tridecane compound. An alternative route to (1R,2S,9S)-11-methyl-7,11-diazatricyclo[7.3.1.02,7]tridecane
starting from the natural product, (Ϫ)-cytisine, is described. This simple three-step route furnished gram-quantities
of a (ϩ)-sparteine surrogate. X-Ray crystallography of an intermediate in the route, (1R,5S,12S)-3-methoxycarbonyl-
decahydro-1,5-methanopyrido[1,2-a][1,5]diazocin-8-one, enabled the stereochemistry of all of the tricyclic diamines
described in this paper to be unequivocally established. Two other diamines, starting from (S)-proline and resolved
2-piperidine ethanol, were prepared using standard methods. These diamines lacked the bispidine framework of
(Ϫ)-sparteine and were found to impart vastly inferior enantioselectivity. It was concluded that, for the asymmetric
lithiation–substitution of N-Boc pyrrolidine, a rigid bispidine framework and only three of the four rings of
(Ϫ)-sparteine are needed for high enantioselectivity. Furthermore, it is shown that diamine (1R,2S,9S)-11-methyl-
7,11-diazatricyclo[7.3.1.02,7]tridecane is the first successful (ϩ)-sparteine surrogate.
Introduction
In 1991, Kerrick and Beak reported the conversion of N-Boc
pyrrolidine 1 into substituted pyrrolidines such as (S)-2 (79–
96% ee) via enantioselective lithiation with sec-butyllithium
and (Ϫ)-sparteine 3 followed by reaction with a range of
electrophiles.1 The first example of high enantioselectivity in
asymmetric lithiation–substitution using sec-butyllithium and
(Ϫ)-sparteine 3 (for the α-functionalisation of O-alkyl carb-
amates) was described by Hoppe et al. in the previous year.2
Since these seminal contributions, a wealth of successful uses
of (Ϫ)-sparteine 3 in asymmetric synthesis has been docu-
mented in reviews3 and many publications.4 The impact of
(Ϫ)-sparteine 3 as a chiral ligand in organolithium chemistry
has been incredible and, more recently, the use of (Ϫ)-sparteine
3 has been extended to magnesium5 and palladium.6 A specific
example of the functionalisation of N-Boc pyrrolidine 1,
reported by the Beak group, is shown in Scheme 1.1,7
Beak’s optimised conditions for the conversion of N-Boc
pyrrolidine 1 into (S)-2 involved treatment of 1 with 1.2
equivalents of sec-butyllithium/(Ϫ)-sparteine 3 in diethyl ether
at Ϫ78 ЊC for 4 h followed by reaction with trimethylsilyl
chloride. In this way, an 87% yield of (S)-2 of 96% ee was
obtained.1,7 It was shown that isopropyllithium could also be
used for the lithiation but other alkyllithiums, other solvents
and the use of sub-stoichiometric amounts of alkyllithum/(Ϫ)-
sparteine 3 gave considerably lower yields and enantio-
Scheme 1
selectivity. Through a very detailed investigation,8,9 Beak and
co-workers demonstrated that the reaction proceeded through
the rapid formation of an alkyllithium–diamine complex and
subsequent relatively slow but highly enantioselective lithiation/
deprotonation (via Boc carbonyl organolithium “delivery”10) to
give a configurationally stable organolithium species that was
trapped by trimethylsilyl chloride. In more recent work, Wiberg
and Bailey have calculated a transition state model11,12 that
accounts for the observed enantioselectivity in the lithiation
of N-Boc pyrrolidine 1. Key steric interactions were identified
to explain the differences in transition state energies for the
removal of the pro-R and pro-S protons.
As the sense of asymmetric induction imparted by (Ϫ)-
sparteine 3 in the α-functionalisation of N-Boc pyrrolidine
1 is opposite to that present in natural (S)-proline, the (Ϫ)-
sparteine-mediated methodology is a useful stereochemical
complement to synthetic routes from (S)-proline. Beak has
demonstrated the usefulness of the N-Boc pyrrolidine method-
ology by preparing amino alcohol 4 (>99.5% ee)13 (precursor
to Corey’s CBS reagent) and the hydrochloride salt of 2,5-di-
methylpyrrolidine 5 (>99% ee)7 (a well-used chiral auxiliary).
Our own group has utilised similar methodology in a new route
to diamine 6 (85% ee)14 whilst Majewski et al. used the reaction
between lithiated N-Boc pyrrolidine and a tartaric acid-derived
† Electronic supplementary information (ESI) available: Experimental
procedures for the synthesis of 34, 37, rac-25, 39, rac-41, (R)-41, 44,
(ϩ)-24 and (S)-48ؒCSA. See http://www.rsc.org/suppdata/ob/b3/
b308410h/
‡ Author to whom correspondence regarding the X-ray crystal structure
should be addressed.
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3977

View Article Online
aldehyde as a key step in the synthesis of 8-epi-1-deoxycastano-
spermine 7.15 Most recently, Dieter et al. have expanded the
synthetic utility of lithiated N-Boc pyrrolidine by transmetal-
lating the organolithium to the organocuprate (with very little
loss of stereochemical integrity) enabling a wider range of
electrophiles to be coupled to the pyrrolidine unit.16 In this
way, vinyl-substituted pyrrolidine 8 was prepared in 86% ee.
Extension of the N-Boc-pyrrolidine methodology to other
five-membered ring systems has also proved very successful:
Beakand co-workers reported the highly enantioselective
lithiation–substitution of N-Boc indoline 9,17 Kise et al.
described high enantioselectivity in the lithiation of N-Boc
oxazolidine 1018 and Coldham et al. developed a new route to
chiral diamines via α-functionalisation of N-Boc imidazoline 11
(at the position indicated by the arrow).19 In contrast, lithiation–
substitution of the corresponding six-membered ring, N-Boc
piperidine 12, using sec-butyllithum-(Ϫ)-sparteine 3 was much
less satisfactory: the trimethylsilyl adduct was generated from
N-Boc piperidine 12 in only 8% yield and with 74% ee.20 This
difference between the five- and six-membered ring systems is
dramatic but is not without precedent and has been rationalised
using computational methods.12,20
Unfortunately, none of the methods described above pro-
vides a general solution to the fact that (ϩ)-sparteine 3 is not
readily accessible. Thus, there is still a need to develop a ligand
that behaves in an enantiocomplementary fashion to (Ϫ)-
sparteine 3. With this in mind, Beak and co-workers investi-
gated a wide range of ligands for the conversion of N-Boc
pyrrolidine 1 into trimethylsilyl adduct 2.36 A detailed overview
of the results is presented in Scheme 2 and provides insight into
some of the key design features that are required in developing
new types of ligand that could function as (ϩ)-sparteine 3
surrogates. The isolated yield (or the GC yield) reflects the
degree of lithiation of N-Boc pyrrolidine 1 and this is strongly
dependent on the steric hindrance of the alkyllithium-chelated
ligand. For example, sec-butyllithium/TMEDA lithiates N-Boc
pyrrolidine 1 in 30 min whereas sec-butyllithium/(Ϫ)-sparteine
3 requires at least 4 h to reach the same degree of lithiation.
With (Ϫ)-sparteine 3, an 87% yield of (S)-2 of 96% ee was
obtained (after 4 h lithiation time). In contrast, use of (Ϫ)-
isosparteine 13 led to only 10% lithiation of 1 under the
same conditions and (S)-2 was obtained in a reduced 61% ee.
Compared to (Ϫ)-sparteine 3, an alkyllithium complexed by
(Ϫ)-isosparteine 13 is more sterically encumbered as both six-
membered rings extend outwards towards the organolithium.
In (Ϫ)-sparteine 3, one of the rings is held away from the
lithium making the resulting organolithium complex more
reactive. Increased steric hindrance around the organolithium
also correlates with lower enantioselectivity. Related results
were also observed in a series of bispidines (14–16). The
best result was obtained with less sterically hindered bispidine
14 (compared to C2-symmetric 16): adduct (R)-2 of 75% ee
was obtained in 51% yield. Although the reaction mixture
was heterogeneous, the result is notable since the sense of
induction with bispidine 14 was opposite to that obtained
with (Ϫ)-sparteine 3. It was also found that replacing the
N-methyl group in 14 with an oxygen atom (as in 15) was
unsatisfactory.
Beak has also studied a series of non-bispidine based ligands.
Diamine 17 was a very good ligand for the reaction but
furnished adduct 2 in essentially racemic form. This has been
rationalised using computational methods.37 Of the (S)-pro-
line-derived ligands 18–23, some interesting trends emerge.
Ligands 19 and 20 (containing an O–Li group, generated
from the corresponding alcohol using an extra equivalent of
alkyllithium) gave the highest enantioselectivity. Interestingly,
changing the configuration at the prolinol group led to a
reversal in the enantioselectivity indicating that this is the most
important chiral centre in the ligands. Use of sec-butyllithum/
ligand 19 generated a very good 63% yield of (S)-2 of 72% ee
(same sense of induction as that obtained with (Ϫ)-sparteine 3).
Increased steric hindrance (as in 21), the presence of an extra
CH2 linker group (as in 22), replacement of the O–Li group by a
methoxy group (as in 23) or removal of the CH2O–Li group
(as in 18) all gave significantly inferior enantioselectivity com-
pared to 19 and 20. In addition, it is of interest to note that
ligand 19 was the optimal one for Coldham’s dynamic thermo-
dynamic resolution of a racemic lithiated pyrrolidine.35 More
recently, Kozlowski and co-workers have investigated the use of
three different 1,5-diaza-cis-decalins for the conversion of 1
into 2 and the highest enantioselectivity obtained was a modest
26% ee of (R)-238 (although these ligands were more successful
in oxidative biaryl coupling reactions39). Based on the results to
date, the best ligands for the conversion of N-Boc pyrrolidine 1
into trimethylsilyl adduct 2 are 14 and 19. However, the results
with these ligands fall around 20% ee lower than the result
obtained with (Ϫ)-sparteine 3.
One of the main limitations of using (Ϫ)-sparteine 3 in
asymmetric synthesis is that it is only commerically available as
its (Ϫ)-antipode and this limitation has attracted our interest
over the last few years. (ϩ)-Sparteine 3, also reported to be a
natural product,21 can be most easily prepared by resolution
of racemic lupanine (isolated from the natural source) and
subsequent deoxygenation22 although a resolution of racemic
sparteine 1 has also been described.23 An elegant, multi-step,
asymmetric synthesis of (ϩ)-sparteine 3 has been recently dis-
closed by Aubé and co-workers24 but this route is unlikely to
provide multi-gram quantities of (ϩ)-3. Due to the lack of
ready availability of (ϩ)-sparteine 3, several groups have
developed ingenious strategies for synthesising either enantio-
mer of a particular compound using (Ϫ)-sparteine. These
include: (i) changing the solvent;25 (ii) using electrophiles that
react with retention or inversion of configuration;26 (iii) using a
stereodivergent 1,3-elimination process;27 (iv) using deuterium
as a “blocking group”;28 (v) using secondary and tertiary
amides;29 (vi) making use of a transmetallation protocol;30 (vii)
using different chelating groups;31 and (viii) using a dilithiation
protocol.32 In addition, there are certain reactions which pro-
ceed with higher enantioselectivity using bisoxazoline ligands
(readily available in either enantiomeric form) compared to
(Ϫ)-sparteine 3.33 With specific reference to the synthesis of
enantiomerically enriched pyrrolidines, Gawley et al. have
described a detailed study on the stereochemistry (retention or
inversion) of reaction of a lithiated pyrrolidine with different
electrophiles34 and Coldham et al. have recently described an
approach based on the dynamic thermodynamic resolution of a
racemic organolithium species.35
In this paper, we provide additional information on the effect
of ligand structure on the enantioselectivity of converting 1
into 2 and we also specifically address the issue of developing a
ligand that behaves in an enantiocomplementary fashion to
(Ϫ)-sparteine 3. The new ligands were designed by considering
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3978

View Article Online
Scheme 2
the 3-D structure of (ϩ)-sparteine 3 in its lithium-chelating
conformation and by studying the calculated transition state
for lithiation of N-Boc pyrrolidine 1 by an alkyllithium/(Ϫ)-
sparteine 3 complex11,12 (Scheme 3). It appeared to us that one
of the six-membered rings in (ϩ)-sparteine 3 was held away
from the lithium ion and we speculated that it may not be
required for high enantioselectivity in sparteine reactions.
In this way, we designed (ϩ)-sparteine-like diamines 24 and 25.
On further inspection, we wondered whether the rigid bispidine
framework was required for high enantioselectivity. Hence, we
designed diamines 26 and 27 which could in principle mimic
(ϩ)-sparteine 3. Diamines 26 and 27 are also analogues of
ligand 22 (see Scheme 2), previously studied by Beak. After
considerable effort, we succeeded in synthesising ligands 24, 26
and 27 in enantioenriched forms and were able to test their
efficacy in the conversion N-Boc pyrrolidine 1 into 2.40 Herein
we present our results.
Scheme 4 Reagents and conditions: i, Methyl vinyl ketone, Et2O, 0 ЊC,
1 h; ii, HCl(aq), 100 ЊC, 2.5 h; iii, PtO2, H2, HCl(aq), EtOH, rt, 24 h;
iv, ethyl acrylate, Et3N, EtOH, rt, 66 h; v, (a) NaH, NaOEt, xylene,
reflux, 5 h; (b) HCl(aq), reflux, 16 h; vi, (a) LHMDS, THF, Ϫ78 ЊC, 2 h;
(b) HCl(aq), reflux, 16 h; vii, β-alanine ethyl esterؒHCl, K2CO3, EtOH,
reflux, 60 h; viii, MeNH2, (CH2O)n, AcOH, MeOH, reflux, 16 h;
ix, N2H4ؒH2O, KOH, diethylene glycol, reflux, 2 h.
Scheme 3
For the synthesis of diamine rac-25, amino ketone 30 was
required and we had most success with an approach described
by King.43 In our hands, commercially available 4,4-diethoxy-
butan-1-amine 28 was reacted with freshly distilled methyl vinyl
ketone to give intermediate 29 which was cyclised via an intra-
molecular Mannich reaction in hydrochloric acid (reaction
temperature of 100 ЊC was essential) to give amino ketone
3044 in 39% distilled yield. Despite attempted optimisation
(including use of a more recent protocol described by
Heathcock and co-workers in their route to petrosins A and
B45), we were unable to improve the yield of 30 and the reaction
was even lower yielding on scales >20 mmol.
Results and discussion
Synthesis and resolution of diamines rac-24 and rac-25
In a previous paper,41 we reported an approach to enantio-
merically enriched diamine 25 starting from (S)-proline which
was unsuccessful due to racemisation in a double Mannich
reaction. Thus, our attention turned to the synthesis of racemic
diamines 24 and 25 and an investigation of methods for their
resolution. We have previously synthesised diamine rac-2541
(albeit unintentionally) whilst diamine rac-24 is a known com-
pound, prepared by Scheiber and Nemes.42 Our optimised
methods for the preparation of multi-gram quantities of
diamines rac-24 and rac-25 are summarised in Scheme 4.
The King route is less direct to six-ring amino ketone 34 as
the corresponding 5,5-diethoxypentan-1-amine is not com-
mercially available. Thus, based on literature precedent,46,47 we
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3979

View Article Online
optimised a three-step approach to amino ketone 34 starting
from ethyl-2-pyridyl acetate 31. Hydrogenation of pyridine 31
under acidic conditions46 gave crude piperidine 32 which was
reacted with ethyl acrylate in the presence of triethylamine48 to
give bis-ester 33 in 69% yield (after distillation) over the two
steps. The Dieckmann condensation of 33 to 3444,47 proceeded
in 65% yield using Leonard et al.’s procedure47 (sodium
hydride/sodium ethoxide in refluxing xylene) but in 87% yield
using lithium hexamethyldisilazide (LHMDS) as the base.
We prefer the LHMDS procedure and this also scaled up
more reliably than Leonard’s protocol. In both methods, it is
necessary to remove all of the THF (by evaporation) before
refluxing overnight in hydrochloric acid. An alternative
approach to bis-ester 33 starting from iodo ester 35 (prepared
according to the literature route49–51) was also briefly investi-
gated. Based on Bunce et al.’s approach49 to cyclic amines,
iodo ester 35 and β-alanine ethyl ester hydrochloride underwent
an intermolecular substitution reaction followed by an intra-
molecular conjugate addition in the presence of base to give
bis-ester 33. We found that the use of potassium carbonate
in place of triethylamine gave improved yields in this type of
reaction and bis-ester 33 was isolated in 78% yield. However,
the need for the preparation of iodo ester 35 means that we
prefer the three-step route to amino ketone 34 starting from
pyridine 31.
one example. In 1983, Toda et al. reported the resolution of a
series of acetylinic alcohols (e.g. 40 and 41) via inclusion
complex formation with (Ϫ)-sparteine 3.52 Our plan was to use
enantioenriched acetylinic alcohols such as 40 and 41 (resolved
using (Ϫ)-sparteine 3) to resolve racemic diamines 24 and 25
into their respective enantiomers. As diamines 24 and 25 were
significantly “sparteine-like” in structure, we hoped that some
success would be obtained in this approach. Our results are
summarised in Scheme 5.
᎐
Scheme 5 Reagents and conditions: i, HC᎐CMgBr, THF, reflux, 16 h;
᎐
The key step in the route to diamines rac-24 and rac-25 is the
ii, (a) (Ϫ)-sparteine 3, acetone, rt, 16 h; (b) filter to collect crystals;
(c) treat crystals with 2 M HCl(aq); iii, alcohol (R)-40 (>99% ee),
acetone, rt, 16 h; (b) filter to collect crystals; (c) treat crystals with 2 M
double Mannich reaction (eg 34
36), precedented in the
work of Scheiber and Nemes42 as it is here that the required
relative stereochemistry is established. For the double Mannich
reactions of amino ketones 34 and 30, we discovered that a
simple procedure described by Beak and co-workers.36 gave
consistently better results than the more complicated protocol
reported by Scheiber and Nemes.42 Thus, amino ketones 34
and 30 were reacted with methylamine, paraformaldehyde and
acetic acid in refluxing methanol to give diazatricyclic ketones
36 (58% yield) and 37 (29% yield), respectively. These com-
pounds were formed as the only isolable diastereomers and
their relative stereochemistry was assigned as that depicted in
Scheme 4. This has subsequently been unequivocally verified by
X-ray crystallography of an intermediate in an independent
synthesis of diamine (ϩ)-24 (vide infra). As depicted for the
synthesis of 36 (Fig. 1), the observed double Mannich stereo-
chemistry is a consequence of the preferred chair–chair con-
formation of amino ketone 34 and the fact that the second
intramolecular Mannich reaction to form the new ring can only
occur via an axially oriented iminium ion. Finally, standard
Wolff–Kishner reduction removed the carbonyl group in each
of 36 and 37 to generate diamines rac-24 and rac-25. In
our opinion, the best ways of synthesising diamines rac-24 and
rac-25 are presented in Scheme 4: starting from commercial
materials, diamine rac-25 can be prepared in a convenient three-
step route whereas diamine rac-24 is synthesised via a slightly
longer five-step route.
HCl(aq)
.
Racemic acetylinic alcohols 40 and 41 were prepared by
reaction of commercially available ketone 38 and ketone 3953
(prepared by Friedel–Crafts acylation of benzene) with
ethynylmagnesium bromide in refluxing THF (Scheme 5). The
resolution of alcohols rac-40 and rac-41 was accomplished
satisfactorily using Toda’s method.52 As an example, a descrip-
tion of the resolution of rac-40 using (Ϫ)-sparteine 3 is
provided. An equimolar mixture of alcohol rac-40 and (Ϫ)-
sparteine 3 was left for 16 h in acetone. During this period,
some of the solvent evaporated slowly and crystals of an inclu-
sion complex formed (aided by the addition of light petroleum
40–60 ЊC). The crystals were collected by filtration, treated
with 2 M hydrochloric acid and the resulting enantioenriched
alcohol (R)-40 {35% yield, [α]D Ϫ115.2 (c 0.6 in MeOH)} was
isolated after extraction into diethyl ether. This alcohol was
then recomplexed with fresh (Ϫ)-sparteine 3 and the process
repeated to give alcohol (R)-40 {[α]D Ϫ139.6 (c 0.6 in MeOH),
lit.,54 [α]D Ϫ129.0 (c 0.2 in MeOH)} of >99% ee as judged by
chiral HPLC. The yield of alcohol (R)-40 was 28% from rac-40
(out of a maximum 50%). From the first complexation with
(Ϫ)-sparteine 3, a 64% yield of (S)-40 {[α]D ϩ58.4 (c 0.6 in
MeOH)} was also obtained. However, it was not possible to
improve the enantiomeric excess of alcohol (S)-40 beyond
∼50% by further complexation with (Ϫ)-sparteine 3. The
resolution of alcohol rac-41 required three complexations to
reach >99% ee: alcohol (R)-41 {[α]D Ϫ123.6 (c 1.1 in MeOH),
lit.,54 Ϫ114 (c 0.2 in MeOH)} of >99% ee by chiral HPLC was
obtained in 30% yield.
Diamine rac-24 was then partially resolved via inclusion
complex formation with resolved alcohol (R)-40 (>99% ee)
using essentially the same procedure as that described above.
After one complexation, we obtained a 23% yield of diamine
(Ϫ)-24 {[α]D Ϫ15.7 (c 0.5 in EtOH)} of 60% ee and a 68% yield
of diamine (ϩ)-24 {[α]D ϩ6.1 (c 0.6 in EtOH)} of 30% ee. The
enantiomeric excess of these diamines was determined using 1H
NMR spectroscopy in the presence of the chiral shift reagent,
(R)-2,2,2-trifluoro-1-(9-anthryl)ethanol. Initially, we assigned
the absolute stereochemistry of diamine (Ϫ)-24 to be that
depicted in Scheme 5 (i.e. the same as naturally occurring
(Ϫ)-sparteine 3) based on the fact that it was generated from
Fig. 1
With multi-gram quantities of racemic 24 and 25 in hand,
their resolution was investigated next. Classical resolution
methods using tartaric acid (and derivatives), malic acid and
camphorsulfonic acid in a range of solvents met with uniform
failure. In the few cases where crystal formation was observed,
surprisingly, only racemic crystals were isolated. Thus, we
sought substrates where (Ϫ)-sparteine 3 had been successfully
used as a resolving agent and our literature survey revealed only
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3980

View Article Online
the crystals obtained from rac-24 and alcohol (R)-40 (itself
generated from the crystals obtained from rac-40 and (Ϫ)-
sparteine 3). This was subsequently confirmed when we syn-
thesised diamine (ϩ)-24 starting from (Ϫ)-cytisine 42 of known
absolute stereochemistry (vide infra). Presumably, diamine (Ϫ)-
24 of 60% ee could be further enriched with additional com-
plexations. However, as supply of alcohol (R)-40 and diamine
rac-40 was limited, complexation a second time with (Ϫ)-
sparteine was not attempted. The resolution was reproducible
in the 50–60% ee range and after repeating it a few times, we
obtained sufficient quantity of diamine (Ϫ)-24 of 55% ee for its
evaluation as a (Ϫ)-sparteine 3 mimic. In contrast to diamine
rac-24, the five-ring analogue, diamine rac-25 was completely
resistant to resolution using alcohol (R)-40 (no crystals were
formed). Even alcohol (R)-41, which was successfully used to
resolve racemic sparteine,52 did not generate any crystals when
combined with rac-25. As Toda et al. found, small changes to
the structure of the alcohol or the diamine can affect inclusion
complex formation considerably.52,54
Scheme 6 Reagents and conditions: i, (a) NH4OH(aq), CH2Cl2, MeOH,
rt, 3 days; (b) 3 M HCl(aq); (c) NH4OH(aq) then extract into CH2Cl2;
(d) recrystallise from acetone; ii, Et3N, MeO2CCl, CH2Cl2, rt, 3.5 h;
iii, PtO2, H2, EtOH, rt, 5 h (typically); iv, LiAlH4, THF, reflux, 16 h.
In summary, diamine (Ϫ)-24 of 55% ee was generated via
resolution of the racemate (prepared in five steps) using alcohol
(R)-40 (itself obtained by resolution using (Ϫ)-sparteine 3).
Although this is a fairly inefficient route to diamine (Ϫ)-24
that would only function as a mimic of (Ϫ)-sparteine 3, it
does represent the first synthesis of non-racemic diamine 24.
Unfortunately, attempts to prepare diamine (ϩ)-24 via this
approach (from either the mother-liquors of the resolution
of alcohol rac-40/41 or the mother-liquors of the resolution of
diamine rac-24) never exceeded 30% ee. In addition, the
approach could not be extended to prepare enantioenriched
five-ring-substituted diamine 25.
Synthesis of diamine (؉)-24 from (؊)-cytisine
Fig. 2 Chem3-D® representation of the X-ray crystal structure of
lactam 44.
Due to the limited success in the resolution of diamines rac-24
and rac-25, we sought an alternative route to the enantio-
enriched diamines and were attracted to the possibility of
starting from (Ϫ)-cytisine 42, another natural product from the
same family as (Ϫ)-sparteine 3. Unfortunately, (Ϫ)-cytisine 42
is prohibitively expensive from commercial suppliers.55 How-
ever, full details of the extraction of (Ϫ)-cytisine 42 from
Laburnum anagyroides seeds have been recently reported in the
supporting information of a communication from Lasne
and co-workers that describes the preparation of (Ϫ)-cytisine
analogues for a study of nicotinic receptors.56 We found this
extraction procedure to be a convenient source of multi-gram
quantities of (Ϫ)-cytisine 42 suitable for the preparation of
diamine (ϩ)-24 in >90% ee. An overview of the straightforward
synthetic route is shown in Scheme 6.
occurred exclusively on the less sterically hindered exo face
of pyridone 43, as precedented in the work of Ohmiya and
co-workers.61
Finally, reduction of lactam 44 with lithium aluminium
hydride (in THF at reflux) converted the methyl carbamate
protecting group into the required N-methyl substituent as well
as transforming the lactam into the piperidine ring (Scheme 6).
In this way, diamine (ϩ)-24 {[α]D ϩ26.5 (c 1.0 in EtOH)} of
>90% ee (determined using 1H NMR spectroscopy in the
presence of the chiral shift reagent, (R)-2,2,2-trifluoro-1-(9-
anthryl)ethanol) was obtained in 88% yield. As far as can be
1
judged by H and 13C NMR spectroscopy, diamine (ϩ)-24 is
Following the literature protocol,56 (Ϫ)-cytisine 4257,58
was extracted in 1.1% mass yield (after crystallisation from
acetone) from 200 g of powdered Laburnum anagyroides seeds
(Vilmorin, France). Next, the free amine was N-protected using
10 equivalents of each of triethylamine and methyl chloro-
formate to give methyl carbamate 43 in 92% yield.59 The
literature conditions reported for this step utilised the large
excess of the reagents but we found that 1.1 equivalents of
each worked equally well (see Experimental section for
details). Hydrogenation of the pyridone functionality using
platinum() oxide (Adams’ catalyst)60,61 was then explored.
Reaction of pyridone 43 in ethanol with hydrogen (at
atmospheric pressure) and Adams’ catalyst furnished lactam
44 in 82% yield after chromatography. Lactam 44 was the only
diastereomerically pure (as we have previously prepared its
diastereomer62 which has distinguishable NMR spectroscopic
data) and the absolute stereochemistry of (ϩ)-24 was assigned
based on the known absolute stereochemistry of (Ϫ)-cytisine 42.
Further optimisation of the three-step route from recrystal-
lised (Ϫ)-cytisine 42 was carried out and we have two altern-
ative protocols for preparing diamine (ϩ)-24. Chromatography
can be avoided throughout the synthesis and crude methyl
carbamate 43 can be hydrogenated to give lactam 44 in 99%
crude yield. However, subsequent lithium aluminium hydride
reduction of crude 44 generated in this way was lower yielding.
Thus, we recommend purification of carbamate 43 using
chromatography to give >90% yield of high purity 43.
Then, pyridone hydrogenation of pure 43 followed by lithium
aluminium hydride reduction (without purification of the
intermediate 44) then gave diamine (ϩ)-24 in a reproducible
86% yield after purification by distillation.
diastereomer produced in this reaction (according to 1H and 13
C
NMR spectroscopy of the crude product) and its stereo-
chemistry was unequivocally assigned as that indicated by
X-ray crystallography (Fig. 2). As shown in Fig. 2, H-12 is
cis to the methylene bridge between carbons C-1 and C-5
and, crucially, this is the relative stereochemistry required in
sparteine-like diamine 24 (see Scheme 3). Thus, hydrogenation
To summarise, the route to diamine (ϩ)-24 starting from
recrystallised (Ϫ)-cytisine (Scheme 6) has several notable
features. Firstly, the absolute stereochemistry of (Ϫ)-cytisine is
suitable for generating a mimic of (ϩ)-sparteine 3. Secondly,
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3981

View Article Online
the pyridone hydrogenation affords the relative stereochemistry
required in diamine (ϩ)-24 with complete control. Thirdly,
the methyl carbamate group acts as both an N-protecting
group and a masked N-methyl group (revealed by a late-stage
lithium aluminium hydride reduction). And, finally, the simple,
three-step route is high yielding (79% overall) and produces
gram quantities of diamine (ϩ)-24.
diamines containing one less methylene group in the side
chain.69 Standard mesylation of (S)-49 was followed by heating
the crude mesylate in toluene with added piperidine and
DBU (according to the literature procedure for a similar
compound69). After purification by chromatography, amino
carbamate (S)-50 was isolated in 66% yield. Subsequent reduc-
tion using lithium aluminium hydride at room temperature gave
diamine (S)-26 in 89% yield after distillation. Diamine (S)-26
was of 90% ee as shown by 1H NMR spectroscopy in the
presence of the chiral shift reagent (R)-2,2,2-trifluoro-1-(9-
anthryl)ethanol). Essentially the same reactions were used to
convert Boc-protected amino alcohol (S)-47 into amino
carbamate (S)-51 and thence into diamine (S)-27 (presumed
>95% ee). However, it was found that DBU was not required
in the substitution step (76% yield of (S)-51 in the absence of
DBU) and the lithium aluminium hydride reduction step
was carried out at reflux (for reduction of the more sterically
hindered Boc group).
Synthesis of diamines (S )-26 and (S )-27
The preparation of enantioenriched diamines (S)-26 and (S)-
27, lacking the rigid bispidine framework, has also been carried
out. The (S) absolute configuration was selected as diamines
with this configuration were expected to behave as (ϩ)-
sparteine mimics (see Scheme 3). Diamine (S)-26 was prepared
from the known63,64 camphorsulfonate salt of piperidine
ethanol (S)-48ؒCSA and diamine (S)-27 was prepared from
known41 protected amino ester (S)-45, ultimately derived from
(S)-proline. The full synthetic routes are depicted in Scheme 7.
Evaluation of the new diamines as sparteine surrogates
With new, enentioenriched diamines in hand, the final objective
was to evaluate them as sparteine surrogates in the lithiation–
substitution of N-Boc pyrrolidine 1. As a benchmark for direct
comparison, we intitially repeated Beak’s conversion of N-Boc
pyrrolidine 1 into trimethylsilyl adduct (S)-2 using 1.3 equiv-
alents of sec-butyllithium and (Ϫ)-sparteine 3. In our hands,
an 87% yield of (S)-2 of 90% ee (determined by chiral GC) was
obtained (Scheme 8).
Scheme 7 Reagents and conditions: i, LiOH(aq), dioxane, rt, 16 h; ii,
BH3ؒMe2S, THF, reflux, 1 h iii, (a) K2CO3(aq), EtO2CCl, CH2Cl2, rt, 20
h; (b) K2CO3, MeOH, rt, 16 h; iv, (a) Et3N, MsCl, CH2Cl2, 0 ЊC then rt
for 1 h; (b) piperidine, toluene, DBU (for 49 only), reflux, 16–24 h;
v, LiAlH4, THF, rt for 48 h (for 26) or reflux for 16 h (for 27) (see
Experimental section for details).
The preparation of Boc-protected amino ester (S)-45 using
an Arndt–Eistert homologation method has previously been
described by us.41 It was converted in good yield (86% overall)
into Boc-protected amino alcohol (S)-4765 via acid (S)-4666 by
ester hydrolysis and borane reduction.67
Commercially available 2-piperidine ethanol rac-48 was
resolved using (ϩ)-camphorsulfonic acid to give a mediocre
12% yield of (S)-48ؒCSA {[α]D ϩ28.1 (c 1.1 in CHCl3) (lit.,63
ϩ34.0 (c 1.0 in CHCl3)} of approximately 90% ee (based on the
% ee determined for diamine (S)-26, vide infra). The conversion
of salt (S)-48ؒCSA into ethyl carbamate-protected amino
alcohol (S)-49 was best achieved by simultaneous N- and O-
protection using ethyl chloroformate and then subsequent
deprotection of the more labile carbonate group. Thus, treat-
ment of salt (S)-48ؒCSA with three equivalents of ethyl chloro-
formate in aqueous potassium carbonate–dichloromethane
(based on related literature protocols68) afforded the N-
Scheme 8
The same reaction was repeated using diamines (Ϫ)-24 (55%
ee), (ϩ)-24 (>90% ee), (S)-26 (90% ee) and (S)-27 (>95% ee
assumed). In addition, as we were unable to obtain enantio-
enriched diamine 25, we evaluated its efficacy as a ligand by
carrying out the conversion of 1 into rac-2 using diamine
rac-25. The results are presented in Scheme 8 together with
the result reported by Lesma, Silvani and co-workers70 using
diamine 52 (which is equivalent to the BCD rings of (Ϫ)-
sparteine 3 as depicted in Scheme 8). Diamine rac-25 did
1
and O-protected amino alcohol (characterised by H and 13C
NMR spectroscopy). This crude product was then reacted with
potassium carbonate in methanol to generate ethyl carbamate-
protected amino alcohol (S)-49 in 95% yield.
For the introduction of the second amino group, we planned
to activate the hydroxyl groups (e.g. as a tosylate or mesylate)
in (S)-49 and (S)-47 and displace them with piperidine. A
similar route was used by Hendrie and Leonard to prepare
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3982

View Article Online
support lithiation of N-Boc pyrrolidine 1 as judged by the
55% yield of rac-2 obtained. Partially resolved diamine (Ϫ)-24
(55% ee) generated (S)-2 of 53% ee (by chiral GC) indicating
that the A-ring of (Ϫ)-sparteine 3 is important for high enantio-
selectivity. This result was subsequently confirmed by Lesma,
ment of cyclooctene oxide and oxidative kinetic resolutions of
alcohols).40 Future work will be carried out in assessing the full
scope and usefulness of diamines such as (ϩ)-24 derived from
(Ϫ)-cytisine and in studying the effect of the N-substituent
in (Ϫ)-cytisine-derived diamines on the enantioselectivity of
various reactions/processes. These results will be reported in
due course.
Silvani and co-workers using diamine (Ϫ)-24 of >98% ee.70
A
comparison with diamine 52 is useful at this point: diamine 52
gave (S)-2 (same sense of induction as with (Ϫ)-sparteine 3)
in only 21% ee. This clearly indicates that the ABC rings of
(Ϫ)-sparteine 3 are needed for high enantioselectivity. This was
our prediction based on Chem3-D® modelling of (Ϫ)- and (ϩ)-
sparteine 3 in its lithium-chelating conformation and sub-
sequently by examination of Wiberg and Bailey’s calculated
transition state model for lithiation of N-Boc pyrrolidine 1
using (Ϫ)-sparteine 3.11,12 Of most significance, use of the easily
prepared diamine (ϩ)-24 gave an 84% yield of (R)-2 of 90% ee
(by chiral GC). This is the first example of a diamine that
behaves in an enantiocomplementary fashion to (Ϫ)-sparteine
3 and diamine (ϩ)-24 is also successful as a (ϩ)-sparteine
surrogate in other applications.40
Unfortunately, the more conformationally flexible diamines
(S)-26 and (S)-27 did not lead to any enantioselectivity in the
lithiation of N-Boc pyrrolidine 1. Essentially racemic adducts 2
(as judged by very low optical rotation data) were generated
in moderate yields from these reactions. We had hoped that
diamines (S)-26 and (S)-27 would self-assemble around the
lithium counterion to provide a similar chiral architecture to
that found in the alkyllithium-(Ϫ)–sparteine 3 complex. How-
ever, our reasoning in this case was at odds with the experi-
mental results. A comparison with the results obtained in
Beak’s detailed study (see Scheme 2) shows that the structurally
most similar diamines (18 and 22) gave a maximum of 20% ee.
Our results confirm the importance of the rigid bispidine
framework and that Beak’s diamine 19 (containing an O–Li
group) remains the best non-bispidine-based ligand reported
thus far for the asymmetric lithiation–substitution of N-Boc
pyrrolidine 1.
Experimental
General
Water is distilled water. Et2O and THF were freshly distilled
from sodium benzophenone ketyl whereas CH2Cl2 was freshly
distilled from calcium hydride. Triethylamine was stored over
potassium hydroxide pellets. sec-Butyllithium was titrated
against N-benzylbenzamide before use.72 All non-aqueous
reactions were carried out under oxygen-free nitrogen using
oven-dried glassware. Light petroleum refers to the fraction
boiling in the range 40–60 ЊC. Brine refers to a saturated
aqueous solution of sodium chloride. Flash column chromato-
graphy was carried out using ICN Biomedicals GmbH silica
(60 Å) or Fisher Matrex silica 60, 70–200 micron. Thin layer
chromatography was carried out using commercially available
Merck F254 aluminium-backed silica plates. For Kugelrohr dis-
tillations, the temperatures quoted correspond to the oven
temperatures.
Proton (270 MHz or 400 MHz) and carbon (67.9 or 100.6
MHz) NMR spectra were recorded on a JEOL EX-270 or a
JEOL ECX-400 instrument using an internal deuterium lock.
All samples were recorded as solutions in CDCl3 and chemical
shifts are quoted in parts per million downfield of tetramethyl-
silane. Carbon NMR spectra were recorded with broad band
proton decoupling and were assigned using DEPT experiments.
Melting points were measured on a Gallenkamp melting
point apparatus and are uncorrected. Microanalyses were
carried out at the University of Newcastle on a Carlo Erba
1106 elemental analyser and weighed using a Mettler MT 5
microbalance. Infrared spectra were recorded on an ATI
Matteson Genesis FT-IR spectrometer. Chemical ionisation
and high resolution mass spectra were recorded on a Fisons
Analytical (VG) Autospec spectrometer. Optical rotations were
recorded at 20 ЊC on a Jasco DIP-370 polarimeter (using the
Conclusion
In this paper, optimised routes for the preparation of sparteine-
like diamines rac-24 and rac-25 have been described. The
methods investigated for resolution of these diamines were of
limited success. However, recent results from the Hodgson
group have demonstrated some examples where racemic
diamines 24 and 25 outperform (Ϫ)-sparteine in terms of chem-
ical yield.71 This may be due to the fact that diamines 24 and
25 are less sterically hindered than (Ϫ)-sparteine 3. Thus, it is
likely that racemic diamines 24 and 25 could have other useful
applications as modified TMEDA ligands.
The results presented in this paper, together with those
recently reported by Lesma, Silvani and co-workers70 provide
significant information on exactly which parts of the (Ϫ)-
sparteine 3 structure are responsible for high enantioselectivity
in the enantioselective lithiation of N-Boc pyrrolidine. It
appears that the rigidity imparted by the bispidine framework
is crucial. In addition, we conclude that the ABC rings of
(Ϫ)-sparteine 3 (as drawn in Scheme 8) are the key structural
features necessary for high enantioselectivity. The D-ring of
(Ϫ)-sparteine is superfluous. On inspection, the results appear
to be consistent with the calculated transition state model for
lithiation of N-Boc pyrrolidine 1 reported by Wiberg and
Bailey.11,12
sodium D line; 589 nm) and [α]D are given in 10Ϫ1 deg cm2 gϪ1
.
Ethyl (E)-7-iodohept-2-enoate 3549–51 and methyl ester (S)-
4541 were prepared according to the literature routes.
Hexahydroindolizin-7(1H)-one 30
Freshly distilled methyl vinyl ketone (2.1 cm3, 25.7 mmol) was
added dropwise to a stirred solution of 4,4-diethoxybutan-1-
amine 28 (3.8 cm3 of 90% technical grade quality, 19.8 mmol)
in Et2O (10 cm3) at 0 ЊC under nitrogen. After stirring for 1 h at
0 ЊC, 2.5 M hydrochloric acid (50 cm3) was added and the
mixture was warmed to room temperature. The two layers were
separated and the aqueous layer was heated at reflux (approx.
100 ЊC) for 2.5 h without a reflux condenser. Then, after cooling
to room temperature, the aqueous mixture was evaporated
under reduced pressure and saturated aqueous potassium
carbonate solution (100 cm3) was added. The mixture was
extracted with CHCl3 (3 × 100 cm3) and the combined CHCl3
extracts were dried (MgSO4) and evaporated under reduced
pressure to give the crude product. Purification by Kugelrohr
distillation gave amino ketone 30 (1.1 g, 39%) as a colourless
oil, bp 60–65 ЊC/1.0 mmHg (lit.,40 60–63 ЊC/1.0 mmHg); RF
(9 : 1 CHCl3–MeOH) 0.35. Spectroscopic data identical to that
reported in the literature.41
Finally, the simple, three-step synthesis of diamine (ϩ)-24 of
>90% ee starting from (Ϫ)-cytisine 42 is a major development
from this project. Diamine (ϩ)-24 is the first reported (ϩ)-
sparteine surrogate and, as well as the lithiation of N-Boc
pyrrolidine reported in this paper, we have demonstrated that it
is enantiocomplentary to (Ϫ)-sparteine 3 in other applications
(e.g. α-functionalisation of O-alkyl carbamates, rearrange-
Ethyl 3-[2-(2-ethoxy-2-oxoethyl)piperidin-1-yl)propanoate 33
A suspension of platinum() oxide (370 mg, 1.0 mmol), ethyl-
2-pyridyl acetate 31 (5.0 cm3, 32.8 mmol) and 6 M hydrochloric
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3983

View Article Online
acid (8.75 cm3) in EtOH (50 cm3) was stirred at room
temperature under a hydrogen atmosphere (balloon) for 24 h
(or until the mixture became colourless). The solvent was
evaporated to give an off-white solid. 2 M Ammonium
hydroxide solution (50 cm3) was added and the mixture
extracted with Et2O (6 × 25 cm3). The combined Et2O extracts
were dried (MgSO4) and evaporated under reduced pressure to
give crude amino ester 32 (4.2 g, 75%) as a colourless oil. A
solution of this crude amino ester 32, ethyl acrylate (7.85 cm3,
72.4 mmol) and triethylamine (20.2 cm3, 144.9 mmol) in EtOH
(100 cm3) was stirred at room temperature under nitrogen for
66 h. The solvent was evaporated under reduced pressure to
give the crude product. Purification by Kugelrohr distillation
gave bis-ester 33 (6.0 g, 92%, 69% from 31) as a colourless oil,
bp 175–180 ЊC/1.0 mmHg; RF (97 : 3 CH2Cl2–MeOH) 0.3;
added and the resulting solution was stirred and heated at reflux
for 16 h. After cooling to room temperature, the solution was
carefully neutralised with solid potassium carbonate until
gas evolution stopped and the solution was saturated. The
solid was removed by filtration and washed with Et2O (15 cm3).
The aqueous filtrate was extracted with Et2O (3 × 10 cm3). The
combined Et2O extracts were dried (Na2SO4) and evaporated
under reduced pressure to give the crude amino ketone 34 (960
mg, 87%) as a pale yellow oil, RF (97 : 3 CH2Cl2–MeOH) 0.35;
identical spectroscopically to that reported in the literature,44
and sufficiently pure for use in the next step.
(1S*,2R*,9R*)-11-Methyl-7,11-diazatricyclo[7.3.1.02,7]tri-
decan-13-one 36
Methylamine (3.25 cm3 of a 2.0 M solution in MeOH,
6.6 mmol) was added dropwise to a stirred solution of amino
ketone 34 (1.0 g, 6.6 mmol), paraformaldehyde (594 mg,
19.9 mmol) and acetic acid (0.4 cm3) in MeOH (7.7 cm3) at
room temperature under nitrogen. The resulting solution was
heated at reflux for 16 h. After cooling to room temperature,
the solvent was evaporated under reduced pressure and 50%
aqueous potassium hydroxide solution (30 cm3) was added to
the residue. The aqeuous mixture was extracted with Et2O (3 ×
60 cm3) and the combined Et2O extracts were dried (Na2SO4)
and evaporated under reduced pressure to give the crude prod-
uct. Purification by Kugelrohr distillation gave diazatricyclic
ketone 36 (803 mg, 58%) as a colourless oil, bp 140–150 ЊC/0.8
mmHg (lit.,42 120–131 ЊC/0.1–0.4 mmHg); RF (4 : 1 CH2Cl2–
MeOH) 0.2; νmax(CHCl3)/cmϪ1 2942 and 1717 (C᎐O); δ (270
νmax(CHCl3)/cmϪ1 2939 and 1726 (C᎐O); δH (270 MHz; CDCl3)
᎐
4.12 (4 H, q, J 7.0, OCH2), 2.93–2.59 (5 H, m, CHN), 2.45 (2 H,
t, J 7.0, CH2CO2), 2.39–2.27 (2 H, m), 1.75–1.28 (6 H, m) and
1.25 (6 H, t, J 7.0, Me); δC (67.9 MHz; CDCl ) 172.6 (C᎐O),
᎐
3
60.4 (OCH2), 60.3 (OCH2), 56.3 (CHN), 50.3 (CH2N), 49.3
(CH2N), 35.7 (CH2CO2), 31.9 (CH2CO2), 30.7 (CH2), 25.2
(CH2), 22.0 (CH2) and 14.2 (Me) (two signals not resolved);
m/z (CI; NH3) 272 [100%, (M ϩ H)ϩ] [Found: (M ϩ H)ϩ,
272.1861. C14H25NO4 requires M ϩ H, 272.1861].
In a separate experiment, a sample of the intermediate
amino ester 32 was purified by Kugelrohr distillation. Ethyl
piperidin-2-yl acetate 32: colourless oil, bp 85–90 ЊC/1.5
mmHg; RF (4 : 1 CH2Cl2–MeOH) 0.5; νmax(CHCl3)/cmϪ1 3327
(NH), 2934 and 1724 (C᎐O); δ (270 MHz; CDCl3) 4.14 (2 H,
᎐
H
᎐
H
q, J 7.0, OCH2), 3.09–3.01 (1 H, m, CHN), 3.02–2.86 (1 H,
m, CHN), 2.66 (1 H, dt, J 3.0 and 12.0, CHN), 2.37–2.33
(2 H, m), 2.11 (1 H, br s, NH), 1.80–1.63 (1 H, m), 1.66–1.51
(2 H, m), 1.49–1.32 (2 H, m) 1.26 (3 H, t, J 7.0, Me) and 1.21–
MHz; CDCl3) 3.32–2.73 (5 H, m), 2.58–2.51 (1 H, m), 2.40–2.17
(2 H, m), 2.24 (3 H, s, NMe), 2.14–1.85 (3 H, m), 1.79–1.50
(4 H, m) and 1.41–1.10 (2 H, m); δC (67.9 MHz; CDCl3) 214.5
(C᎐O), 66.7 (NMe), 62.4 (CH N), 60.4 (CH N), 56.4 (CH N),
᎐
2
2
2
1.12 (1 H, m); δC (67.9 MHz; CDCl ) 172.3 (C᎐O), 60.2
᎐
3
54.9 (CH2N), 52.1 (CHN), 47.4 (CH), 45.4 (CH), 29.9 (CH2),
25.4 (CH2) and 23.4 (CH2); m/z (EI) 208 [28%, Mϩ], 164 (44),
150 (58) and 98 (100) [Found: Mϩ, 208.1578. C12H20N2O
requires M, 208.1576].
(OCH2), 53.2 (CHN), 46.8 (CH2N), 41.6 (CH2), 32.5 (CH2),
26.0 (CH2), 24.5 (CH2) and 14.1 (Me); m/z (CI; NH3) 172
[100%, (M ϩ H)ϩ] [Found: (M ϩ H)ϩ, 172.1341. C9H17NO2
requires M ϩ H, 172.1338].
(1S*,2R*,9R*)-11-Methyl-7,11-diazatricyclo[7.3.1.02,7]tri-
decane or (1S*,5R*,12R*)-3-methyldecahydro-1,5-methano-
pyrido[1,2-a]diazocine rac-24
Ethyl 3-[2-(2-ethoxy-2-oxoethyl)piperidin-1-yl)propanoate 33
A suspension of ethyl (E)-7-iodohept-2-enoate 3549–51 (100 mg,
0.35 mmol), β-alanine ethyl ester hydrochloride (61 mg, 0.39
mmol) and potassium carbonate (147 mg, 1.06 mmol) in EtOH
(2 cm3) was stirred and heated at reflux under nitrogen for 60 h.
After cooling to room temperature, the solvent was evaporated
under reduced pressure and the residue dissolved in Et2O
(10 cm3) and water (10 cm3). The two layers were separated and
the Et2O layer was washed with 5% aqueous sodium thiosulfate
solution (10 cm3) and brine (10 cm3). The combined aqueous
washings were then extracted with Et2O (2 × 10 cm3). The com-
bined Et2O extracts were dried (Na2SO4) and evaporated under
reduced pressure to give the crude product. Purification by flash
column chromatography on silica with CHCl3–MeOH (97 : 3)
as eluent gave bis-ester 33 (75 mg, 78%) as a colourless oil,
identical spectroscopically to that obtained above.
A solution of diazatricyclic ketone 36 (1.3 g, 6.5 mmol), hydra-
zine hydrate (1.7 cm3, 34.1 mmol) and potassium hydroxide
(4.7 g, 84.1 mmol) in diethylene glycol (13 cm3) under nitrogen
was stirred and heated at reflux (230 ЊC) using a silicone oil bath
for 2 h. Then, the volatile components were removed by dis-
tillation for 1 h at 190 ЊC. The reaction mixture was cooled to
room temperature and combined with the previously distilled
fraction. Water (30 cm3) was added and the aqueous mixture
was extracted with Et2O (4 × 10 cm3). The combined Et2O
extracts were washed with 20% aqueous sodium hydroxide
solution (4 × 10 cm3), dried (MgSO4) and evaporated under
reduced pressure to give the crude product. Purification by
Kugelrohr distillation gave diamine rac-24 (850 mg, 68%) as
a colourless oil, bp 150–160 ЊC/0.8 mmHg; RF (4 : 1 CHCl3–
MeOH) 0.1; νmax(CHCl3)/cmϪ1 2937 and 2775; δH (270 MHz;
CDCl3) 3.02–2.82 (4 H, m), 2.23 (1 H, ddd, J 2.0, 3.0 and 11.0),
2.18–2.10 (1 H, m), 2.16 (3 H, s, NMe), 1.90–1.45 (11 H, m)
and 1.38–1.20 (2 H, m); δC (67.9 MHz; CDCl3) 66.4 (CHN),
60.4 (CH2N), 57.6 (CH2N), 56.3 (CH2N), 53.4 (CH2N), 47.4
(NMe), 35.1 (CH), 33.9 (CH2), 30.7 (CH2), 30.5 (CH), 25.6
(CH2) and 25.0 (CH2); m/z (EI) 194 [50%, Mϩ], 164 (44), 150
(58) and 98 (100) [Found: Mϩ, 194.1787. C12H22N2 requires M,
194.1783].
Octahydro-2H-quinolizin-2-one 34
LHMDS (17.7 cm3 of a 1 M solution in THF, 17.7 mmol)
was added dropwise to a stirred solution of bis-ester 33 (2.0 g,
7.4 mmol) in THF (5 cm3) at Ϫ78 ЊC under nitrogen. After
stirring for 2 h at Ϫ78 ЊC, 12 M hydrochloric acid (1.8 cm3) was
added and the solution was warmed to room temperature.
Water (30 cm3) was added and the mixture was extracted with
Et2O (3 × 25 cm3). Then, saturated aqueous potassium
carbonate solution was added to the aqueous layer until pH 10
was obtained. The aqueous layer was extracted with Et2O (3 ×
25 cm3) and the combined Et2O extracts were dried (Na2SO4)
and evaporated under reduced pressure to give the crude β-keto
ester as a yellow oil. Then, 6 M hydrochloric acid (75 cm3) was
1-(2-Chlorophenyl)-1-phenylprop-2-yn-1-ol rac-40
A solution of ortho-chlorobenzophenone 38 (8.0 g, 37.0 mmol)
in THF (10 cm3) was added dropwise to a stirred solution of
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3984

View Article Online
ethynylmagnesium bromide (150 cm3 of a 0.5 M solution in
THF, 75.0 mmol) at room temperature under nitrogen. The
resulting solution was heated at reflux for 16 h. After cooling
to room temperature, the solution was carefully poured into a
mixture of ice (50 cm3), water (150 cm3) and 12 M hydrochloric
acid (5 cm3). Then, the aqueous mixture was extracted with
Et2O (3 × 100 cm3) and the combined Et2O extracts were dried
(MgSO4) and evaporated under reduced pressure to give the
crude product. Purification by flash column chromatography
on silica with light petroleum–EtOAc (6 : 1) as eluent gave
alcohol rac-40 (7.9 g, 88%) as a brown viscous oil, RF (6 : 1 light
1-(9-anthryl)ethanol) as a colourless oil, identical spectro-
scopically to rac-24, [α]D Ϫ15.7 (c 0.5 in EtOH).
To the combined light petroleum washings from above, a
mixture of Et2O (15 cm3) and 2 M hydrochloric acid (10 cm3)
was added. The two layers were separated and the Et2O layer
was extracted with 2 M hydrochloric acid (2 × 10 cm3). To the
combined aqeuous extracts, 20% aqueous sodium hydroxide
solution was added until pH 9. Then, the solution was extracted
with Et2O (3 × 10 cm3) and the combined Et2O extracts were
dried (MgSO4) and evaporated under reduced pressure to
give diamine (ϩ)-24 (503 mg, 68%, 30% ee by 1H NMR spectro-
scopy in the presence of (R)-2,2,2-trifluoro-1-(9-anthryl)-
ethanol) as a colourless oil, identical spectroscopically to rac-24,
[α]D ϩ6.1 (c 0.6 in EtOH).
petroleum–EtOAc) 0.3; νmax(film)/cmϪ1 3546 (OH), 3430 (OH),
᎐
3292 (C᎐CH), 3062, 1656, 1435, 1337, 1180 and 756; δ (270
᎐
H
MHz; CDCl3) 8.05–7.99 (1 H, m), 7.56–7.52 (2 H, m), 7.38–7.25
(6 H, m), 3.30 (1 H, br s) and 2.89 (1 H, s); δC (67.9 MHz;
(1R,5S )-1,2,3,4,5,6-Hexahydro-1,5-methanopyrido[1,2-a][1,5]-
diazocin-8-one or cytisine (؊)-42
CDCl3) 142.7 (ipso-Ar), 140.2 (ipso-Ar), 132.3 (ipso-Ar), 131.2,
᎐
129.4, 128.2, 128.1, 128.0, 126.6, 126.5, 84.2 (C᎐CH), 75.9
᎐
(C᎐CH) and 73.6 (COH); m/z (EI) 242 [21%, 35Mϩ] [Found:
᎐
᎐
A suspension of Cytisus (Laburnum anagyroides) seeds (200 g,
powdered in a food blender, Vilmorin, France) in CH2Cl2
(300 cm3), MeOH (85 cm3) and 35% w/v aqueous ammonium
hydroxide solution (32 cm3) was stirred at room temperature for
67 h. Then, the mixture was filtered and the residue washed
with CH2Cl2 until the filtrate was colourless. The filtrate
was then acidified by dropwise (CAUTION) addition of 3 M
hydrochloric acid until pH 1 was obtained. The two layers were
separated and the aqueous layer was basified by dropwise
(CAUTION) addition of 35% w/v aqueous ammonium
hydroxide solution until pH 11 was obtained. The aqueous
layer was extracted with CH2Cl2 (10 × 30 cm3) and the com-
bined CH2Cl2 extracts were dried (Na2SO4) and evaporated
under reduced pressure to give crude (Ϫ)-cytisine (3.5 g, 1.65%)
as a brown solid. Recrystallisation from acetone gave (Ϫ)-
cytisine 42 (2.3 g, 1.1%, two crops) as a yellow solid, mp 156–
157 ЊC (lit.,57 155 ЊC); [α]D Ϫ67.5 (c 1.0 in CHCl3) (lit.,56 Ϫ76.0
(c 1.0 in CHCl3)). Spectroscopic data identical to that reported
in the literature.58
(35M Ϫ H)ϩ, 241.0420. C15H1035ClO requires M Ϫ H, 241.0420].
(R)-1-(2-Chlorophenyl)-1-phenylprop-2-yn-1-ol 40
(Ϫ)-Sparteine (7.5 cm3, 32.6 mmol) was added dropwise to a
solution of alcohol rac-40 (7.9 g, 32.6 mmol) in acetone
(40 cm3) at room temperature. After allowing some of the
solvent to evaporate over 16 h, colourless crystals formed. Light
petroleum (20 cm3) was added and the crystals were collected
by filtration and washed with light petroleum (3 × 30 cm3).
Then, the crystals were dissolved in a mixture of Et2O (150 cm3)
and 2 M hydrochloric acid (100 cm3). The two layers were
separated and the aqueous layer was extracted with Et2O (2 ×
100 cm3). The combined Et2O extracts were dried (MgSO4)
and evaporated under reduced pressure to give alcohol (R)-40
(2.7 g, 35%) as a brown viscous oil, [α]D Ϫ115.2 (c 0.6 in
MeOH). To the combined light petroleum washes, a mixture of
Et2O (150 cm3) and 2 M hydrochloric acid (100 cm3) were
added. The two layers were separated and the aqueous layer
was extracted with Et2O (2 × 100 cm3). The combined Et2O
extracts were dried (MgSO4) and evaporated under reduced
pressure to give alcohol (S)-40 (5.0 g, 64%) as a brown viscous
oil, [α]D ϩ58.4 (c 0.6 in MeOH).
Using the procedure described above, alcohol (R)-40 {2.7 g,
[α]D Ϫ115.2 (c 0.6 in MeOH), 11.3 mmol} and (Ϫ)-sparteine
(2.6 cm3, 11.3 mmol) in acetone (15 cm3) gave alcohol (R)-40
(2.2 g, 28% from rac-40, >99% ee by chiral HPLC) as a brown
viscous oil, identical spectroscopically to rac-40, [α]D Ϫ139.6
(c 0.6 in MeOH) (lit.,54 Ϫ129.0 (c 0.2 in MeOH)); HPLC:
Chiralcel OD-H, 5% iPrOH in heptane, 1.0 cm3 minϪ1, 215 nm,
10.2 [(R)-40], 11.6 [(S)-40].
(1R,5S )-N-Methoxycarbonyl-1,2,3,4,5,6-hexahydro-1,5-
methanopyrido[1,2-a][1,5]diazocin-8-one 43
Methyl chloroformate (4.6 cm3, 58.3 mmol) was added drop-
wise over 10 min to a stirred solution of (Ϫ)-cytisine 42 (1.11 g,
5.8 mmol) and triethylamine (8.2 cm3, 58.3 mmol) in CH2Cl2
(33 cm3) at 0 ЊC under nitrogen. After stirring for 3.5 h at
room temperature, the solvent was evaporated under reduced
pressure. EtOAc (15 cm3) was added to the residue and the
solids were removed by filtration. The filtrate was evaporated
under reduced pressure to give the crude product. Purification
by flash column chromatography on silica with CH2Cl2–
MeOH–35% w/v aqueous ammonium hydroxide (90 : 9 : 1) as
eluent gave pyridone 43 (1.33 g, 92%) as a colourless gum,
[α]D Ϫ207.0 (c 1.1 in CHCl3) (lit.,59 Ϫ209.0 (c 0.8 in CHCl3)).
Spectroscopic data identical to that reported in the literature.59
(1S,2R,9R)-11-Methyl-7,11-diazatricyclo[7.3.1.02,7]tridecane or
(1S,5R,12R)-3-methyldecahydro-1,5-methanopyrido[1,2-a]-
[1,5]diazocine (؊)-24
(1R,2S,9S )-11-Methyl-7,11-diazatricyclo[7.3.1.02,7]tridecane
or (1R,5S,12S )-3-methyldecahydro-1,5-methanopyrido[1,2-a]-
[1,5]diazocine (؉)-24
A solution of diamine rac-24 (742 mg, 3.8 mmol) in acetone
(3 cm3) was added dropwise to a solution of alcohol (R)-40
(929 mg, 3.8 mmol, >99% ee) in acetone (3 cm3) at room
temperature. After allowing all of the solvent to evaporate over
64 h, pale yellow crystals formed. Light petroleum (5 cm3)
was added and the crystals were collected by filtration and
washed with light petroleum (3 × 5 cm3). Then, the crystals
were dissolved in a mixture of Et2O (15 cm3) and 2 M hydro-
chloric acid (10 cm3). The two layers were separated and the
Et2O layer was extracted with 2 M hydrochloric acid (2 × 10
cm3). To the combined aqeuous extracts, 20% aqueous sodium
hydroxide solution was added until pH 9 was obtained. Then,
the solution was extracted with Et2O (3 × 10 cm3) and the com-
bined Et2O extracts were dried (MgSO4) and evaporated under
reduced pressure to give diamine (Ϫ)-24 (171 mg, 23%, 60% ee
by 1H NMR spectroscopy in the presence of (R)-2,2,2-trifluoro-
A suspension of pyridone 43 (1.37 g, 5.5 mmol) and plati-
num() oxide (130 mg, 0.55 mmol) in MeOH (25 cm3) was
stirred at room temperature under a hydrogen atmosphere
(balloon) for 5 h. The solids were removed by filtration through
Celite and the filter cake was washed with 9 : 1 CH2Cl2–MeOH
(100 cm3). The filtrate was evaporated under reduced pressure
to give the crude product as a white solid. To this crude product
in THF (25 cm3) at 0 ЊC under nitrogen, lithium aluminium
hydride (1.26 g, 33.2 mmol) was added in one portion. The
resulting suspension was stirred and heated at reflux for 16 h.
After cooling to 0 ЊC, Et2O (10 cm3) was added followed by the
portionwise addition (CAUTION) of solid hydrated sodium
sulfate until effervescence ceased. The solids were removed by
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3985

View Article Online
filtration through Celite and the filter cake was washed with
9 : 1 CH2Cl2–MeOH (100 cm3). The filtrate was dried (Na2SO4)
and evaporated under reduced pressure to give the crude
product as a brown oil. Purification by Kugelrohr distillation
gave diamine (ϩ)-24 (920 mg, 86%, >90% ee by 1H NMR
spectroscopy in the presence of (R)-2,2,2-trifluoro-1-(9-
anthryl)ethanol) as a colourless oil, identical spectroscopically
to rac-24, bp 95–110 ЊC/0.8 mmHg; [α]D ϩ26.5 (c 1.0 in EtOH).
99–100 ЊC); [α]D Ϫ30.7 (c 0.6 in CHCl3); [α]D Ϫ37.8 (c 1.4 in
DMF) (lit.,66 Ϫ40.5 (c 1.9 in DMF); RF (1 : 1 light petroleum–
EtOAc) 0.4; νmax(CHCl3)/cmϪ1 1710 (C᎐O, CO H) and 1685
᎐
2
(C᎐O, Boc); δ (270 MHz; CDCl3) 10.26 (1 H, br s, CO2H),
᎐
H
4.24–4.04 (1 H, m, CHN), 3.47–3.23 (2 H, m, CH2N), 3.01–2.81
(1 H, br m), 2.32 (1 H, br dd, J 9.5 and 14.5), 2.16–1.99 (1 H,
m), 1.90–1.69 (3 H, m) and 1.44 (9 H, CMe3); δC (67.9 MHz;
CDCl ) rotamers observed 176.5 (C᎐O, CO H), 154.4 (C᎐O,
᎐
᎐
3
2
Boc), 79.8 (CMe3), 53.8 (CHN), 46.4 and 46.1 (CH2N), 38.8
(CH2), 31.2 and 30.8 (CH2), 28.4 (CMe3) and 22.6 and 22.3
(CH2); m/z (CI; NH3) 230 [24%, (M ϩ H)ϩ], 174 (100), 130 (33)
and 70 (57) [Found: (M ϩ H)ϩ, 230.1396. C11H19NO4 requires
M ϩ H, 230.1392].
(1R,5S,12S )-3-Methoxycarbonyldecahydro-1,5-methano-
pyrido[1,2-a][1,5]diazocin-8-one 44
A suspension of pyridone 43 (4.84 g, 19.5 mmol) and plati-
num() oxide (136 mg, 0.60 mmol) in EtOH (55 cm3) was
stirred at room temperature under a hydrogen atmosphere
(balloon) for 7 days. The solids were removed by filtration
through Celite and the filter cake was washed with 9 : 1
CH2Cl2–MeOH (100 cm3). The filtrate was evaporated under
reduced pressure to give the crude product. Purification by flash
column chromatography on silica with CH2Cl2–MeOH (98 : 2)
(with a few drops of 35% w/v aqueous ammonium hydroxide)
as eluent gave lactam 44 (4.06 g, 82%) as a white solid, mp
121–122 ЊC; [α]D Ϫ180.3 (c 1.0 in CHCl3); RF (97 : 2 : 1 CH2Cl2–
(2S )-tert-Butyl 2-(2-hydroxyethyl)pyrrolidine-1-carboxylate
(S )-47
Borane dimethyl sulfide (5.6 cm3 of a 2 M solution in THF,
11.2 mmol) was added dropwise to a stirred solution of acid
(S)-46 (2.33 g, 10.2 mmol) in THF (50 cm3) at room tem-
perature under nitrogen. The resulting mixture was heated at
reflux for 1 h. After cooling to room temperature, the THF
was evaporated under reduced pressure and the residue was
dissolved in a mixture of CH2Cl2 (60 cm3) and water (20 cm3).
The two layers were separated and the CH2Cl2 layer was washed
with saturated aqueous sodium hydrogencarbonate solution
(20 cm3) and brine (20 cm3), dried (MgSO4) and evaporated
under reduced pressure to give protected amino alcohol (S)-47
(2.22 g, 100%) as a yellow oil, which was sufficiently pure
(by TLC and 1H NMR spectroscopy) for use in the next step.
In a separate experiment, a sample of amino alcohol (S)-47
was purified by flash column chromatography using 2 : 1 light
petroleum–EtOAc as eluent. Amino alcohol (S)-47: pale yellow
oil, [α]D Ϫ13.8 (c 1.1 in CHCl3); RF (2 : 1 light petroleum–
EtOAc) 0.3; νmax(CHCl3)/cmϪ1 1670 (C᎐O); δ (270 MHz;
MeOH–35% w/v aqueous ammonium hydroxide) 0.3; νmax
-
(CHCl3)/cmϪ1 1691 (C᎐O, CO Me) and 1620 (C᎐O, lactam);
᎐
᎐
2
δH (400 MHz; CDCl3) approx. 4 : 1 mixture of rotamers; signals
for major rotamer: 4.78 (1 H, d, J 13.5, CHN), 4.62 (1 H, dd,
J 1.5 and 14.0, CHN), 4.21 (1 H, dd, J 1.5 and 13.5, CHN), 3.61
(3 H, s, Me), 3.47 (1 H, br d, J 10.5, CHN), 3.09 (1 H, br td,
J 2.0 and 13.5, CHN), 2.88 (1 H, dd, J 2.0 and 14.0, CHN), 2.81
(1 H, br d, J 13.5, CHN), 2.48–2.31 (2 H, m), 2.25–2.08 (1 H,
m), 2.02–1.76 (5 H, m) and 1.73–1.55 (2 H, m) [resolved signals
for minor rotamer: 4.38 (1 H, br d, J 13.5, CHN), 4.30 (1 H, br
d, J 13.0, CHN), 3.68 (3 H, s, Me), 3.02 (1 H, dd, J 13.0, CHN)
and 2.96 (1 H, br d, J 14.0, CHN)]; δC (100.6 MHz; CDCl3)
᎐
H
CDCl3) 4.45 (1 H, br s), 4.14–4.00 (1 H, m), 3.65–3.41 (2 H, m),
3.25 (2 H, t, J 6.5, CH2O), 2.01–1.70 (4 H, m), 1.65–1.49 (2 H,
169.8 (C᎐O, CO Me), 156.0 (C᎐O, lactam), 59.4 (CHN), 52.6
᎐
᎐
2
(MeO), 49.0 (CH2N), 45.7 (CH2N), 44.3 (CH2N), 33.3 (CH2),
33.1 (CH), 32.7 (CH2), 27.8 (CH), 27.7 (CH2) and 20.2 (CH2);
m/z (CI; NH3) 253 [100%, (M ϩ H)ϩ] [Found: (M ϩ H)ϩ,
253.1553. C13H20N2O3 requires M ϩ H, 253.1553]; Found: C,
62.0; H, 8.2; N, 11.0%; C13H20N2O3 requires C, 61.88; H, 7.99;
N, 11.10%. Single crystals of compound 44 suitable for X-ray
diffraction were grown from CDCl3.
m) and 1.39 (9 H, CMe ); δ (67.9 MHz; CDCl ) 156.2 (C᎐O),
᎐
3
C
3
79.6 (CMe3), 58.9 (CH2O), 53.4 (CHN), 46.3 (CH2N), 38.1
(CH2), 30.9 (CH2), 28.3 (CMe3) and 23.4 (CH2); m/z (CI; NH3)
216 [37%, (M ϩ H)ϩ], 160 (100), 116 (57) and 70 (32) [Found:
(M ϩ H)ϩ, 216.1605. C11H21NO3 requires M ϩ H, 216.1600].
(2S )-Ethyl 2-(2-hydroxyethyl)piperidine-1-carboxylate (S )-49
Ethyl chloroformate (1.70 cm3, 17.6 mmol) was added dropwise
to a stirred mixture of salt (S)-48ؒCSA (2.12 g, 5.86 mmol) in
CH2Cl2 (100 cm3) and 5% w/v aqueous potassium carbonate
solution (150 cm3) at 0 ЊC. After warming to room temperature,
the resulting mixture was stirred at room temperature for
20 h. Then, CH2Cl2 (50 cm3) and water (100 cm3) were added
and the two layers were separated. The aqueous layer was
extracted with CH2Cl2 (3 × 150 cm3) and the combined CH2Cl2
extracts were dried (MgSO4) and evaporated under reduced
pressure to give the N- and O-diprotected crude product. The
crude product was dissolved in 1% w/v methanolic potassium
carbonate solution (100 cm3) and the resulting solution
was stirred at room temperature for 16 h. The solvent was
evaporated under reduced pressure and water (50 cm3) was
added. The mixture was extracted with CH2Cl2 (3 × 70 cm3)
and the combined CH2Cl2 extracts were dried (MgSO4) and
evaporated under reduced pressure to give protected amino
alcohol (S)-49 (1.12 g, 95%) as a pale yellow oil, RF (5 : 2 light
petroleum–EtOAc) 0.55; νmax(CHCl3)/cmϪ1 3435 (OH), 2945
Crystal structure determination of lactam 44
Crystal data. C13H20N2O3, M = 252.31, orthorhombic, a =
7.9103(4), b = 9.9556(5), c = 15.8702(9) Å, U = 1249.81(11) Å3,
T = 115(2) K, space group P212121, Z = 4, µ(Mo-Kα) = 0.096
mmϪ1, 10029 reflections measured, 3598 unique (Rint = 0.0196)
which were used in all calculations. The final wR(F 2) was 0.0898
(all data).
CCDC reference number 215917.
See http://www.rsc.org/suppdata/ob/b3/b308410h/ for crys-
tallographic data in CIF or other electronic format.
(2S )-tert-Butyl 2-(2-Carboxyethyl)pyrrolidine-1-carboxylate
(S )-46
A solution of lithium hydroxide (11.6 cm3 of a 1.0 M aqueous
solution, 11.6 mmol) and methyl ester (S)-4541 (3.27 g, 13.4
mmol) in dioxane was stirred at room temperature for 16 h.
Water (10 cm3) was added and the mixture was acidified by
dropwise addition of 10% aqueous citric acid solution until
pH 3–4 was obtained. Then, the dioxane as evaporated under
reduced pressure and the aqueous residue was extracted with
EtOAc (3 × 35 cm3). The combined EtOAc extracts were dried
(MgSO4) and evaporated under reduced pressure to give the
crude product. Purification by flash column chromatography
on silica with light petroleum–EtOAc (1 : 1) as eluent gave
acid (S)-46 (2.64 g, 86%) as a white solid, mp 98–100 ЊC (lit.,66
and 1660 (C᎐O); δ (270 MHz; CDCl3) 4.43 (1 H, br s), 4.10
᎐
H
(2 H, q, J 7.0, OCH2Me), 3.97 (1 H, br d, J 12.5), 3.80–3.52
(2 H, m), 3.51–3.30 (1 H, m), 2.70 (1 H, br t, J 13.0), 2.00–1.87
(1 H, m), 1.85–1.34 (7 H, m) and 1.22 (3 H, t, J 7.0, Me);
δC (67.9 MHz; CDCl ) 155.6 (C᎐O), 61.5 (OCH Me), 58.5
᎐
3
2
(CH2OH), 46.5 (CHN), 39.0 (CH2N), 32.2 (CH2), 29.0 (CH2),
25.3 (CH2), 18.9 (CH2) and 14.5 (Me); m/z (CI; NH3) 202
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3986

View Article Online
[100%, (M ϩ H)ϩ] and 156 (30) [Found: (M ϩ H)ϩ, 202.1446.
C10H19NO3 requires M ϩ H, 202.1443].
diamine (S)-26 (656 mg, 89%, 90% ee by 1H NMR spectroscopy
in the presence of (R)-2,2,2-trifluoro-1-(9-anthryl)ethanol) as
a colourless oil, bp 145–150 ЊC/0.6 mmHg; [α]D Ϫ43.0 (c 1.0 in
CHCl3); νmax(CHCl3)/cmϪ1 2937, 2856 and 2790; δH (270 MHz;
CDCl3) 2.74 (1 H, br d, J 13.5), 2.42–2.17 (5 H, m), 2.18 (3 H, s,
NMe). 2.02–1.92 (1 H, m), 1.84–1.77 (1 H, m) and 1.72–1.15
(15 H, m); δC (67.9 MHz; CDCl3) 62.6 (CHN), 57.0 (CH2N),
55.5 (CH2N), 54.7 (CH2N), 43.0 (NMe), 30.8 (CH2), 29.9
(CH2), 29.5 (CH2), 25.8 (CH2), 24.3 (CH2) and 24.2 (CH2);
m/z (CI; NH3) 211 [100%, (M ϩ H)ϩ], 125 (25) and 98 (35)
[Found: (M ϩ H)ϩ, 211.2172. C13H26N2 requires M ϩ H,
211.2174].
(2S )-Ethyl 2-[2-(1-piperidinyl)ethyl]-1-piperidinecarboxylate
(S )-50
Methanesulfonyl chloride (0.58 cm3, 5.63 mmol) was added
dropwise to a stirred solution of protected amino alcohol (S)-
49 (1.12 g, 5.59 mmol) and triethylamine (0.86 cm3, 6.15 mmol)
in CH2Cl2 (20 cm3) at 0 ЊC under nitrogen. After warming
to room temperature, the resulting solution was stirred for
1 h. Then, water (20 cm3) was added and the two layers
were separated. The aqueous layer was extracted with CH2Cl2
(3 × 30 cm3). The combined CH2Cl2 extracts were washed
with saturated aqueous sodium hydrogencarbonate solution
(40 cm3), dried (MgSO4) and evaporated under reduced
pressure to give the crude mesylate. To a stirred solution of the
crude mesylate in toluene (35 cm3), DBU (0.09 cm3, 0.56 mmol)
and piperidine (2.25 cm3, 22.7 mmol) were added. The resulting
solution was heated at reflux for 24 h. After cooling to room
temperature, the solvent was evaporated under reduced
pressure and the residue was purified by flash column chrom-
atography on silica with CH2Cl2–MeOH (9 : 1) as eluent to give
amino carbamate (S)-50 (981 mg, 66%) as a yellow oil,
[α]D Ϫ27.3 (c 1.3 in CHCl3); RF (9 : 1 CHCl3–MeOH) 0.35;
(S )-N-Methyl-[2-(1-piperidinyl)ethyl]pyrrolidine (S )-27
Lithium aluminium hydride (1.01 g, 26.4 mmol) was added
portionwise to a stirred solution of amino carbamate (S)-51
(1.19 g, 4.22 mmol) in THF (45 cm3) at 0 ЊC under nitrogen.
The resulting suspension was heated at reflux for 24 h. After
cooling to room temperature, Et2O (20 cm3) was added followed
by the portionwise addition (CAUTION) of solid hydrated
sodium sulfate until effervescence ceased. The resulting suspen-
sion was then stirred at room temperature for 2 h and the solids
were removed by filtration through Celite. The filtrate was
evaporated under reduced pressure to give the crude product.
Purification by Kugelrohr distillation gave diamine (S)-27
(579 mg, 70%) as a colourless oil, [α]D Ϫ71.5 (c 0.6 in CHCl3);
νmax(CHCl3)/cmϪ1 1230 and 1201; δH (270 MHz; CDCl3) 3.06–
2.97 (1 H, m), 2.45–2.21 (6 H, m), 2.27 (3 H, s, NMe), 2.08 (1 H,
q, J 9.0), 1.99–1.80 (3 H, m), 1.76–1.51 (6 H, m) and 1.49–1.32
(4 H, m); δC (67.9 MHz; CDCl3) 63.3 (CHN), 57.4 (CH2N), 56.9
(CH2N), 54.9 (CH2N), 40.6 (NMe), 31.4 (CH2), 31.0 (CH2),
26.2 (CH2), 24.7 (CH2) and 22.1 (CH2); m/z (CI; NH3) 197
[100%, (M ϩ H)ϩ], 98 (33) and 84 (32) [Found: (M ϩ H)ϩ,
197.2019. C12H24N2 requires M ϩ H, 197.2018].
νmax(CHCl3)/cmϪ1 2941 and 1675 (C᎐O); δH (270 MHz; CDCl3)
᎐
4.32–4.22 (1 H, m), 4.18–3.92 (3 H, m), 2.91–2.78 (1 H, m),
2.76–2.35 (6 H, m), 2.21–2.01 (1 H, m), 1.72–1.25 (13 H, m)
and 1.21 (3 H, t, J 7.0, Me); δ (67.9 MHz; CDCl ) 155.7 (C᎐O),
᎐
C
3
61.1 (CH2O), 55.9 (CH2N), 54.3 (CH2N), 49.0 (CHN), 38.9
(CH2N), 28.8 (CH2), 25.3 (CH2), 24.5 (CH2), 23.3 (CH2), 18.9
(CH2) and 14.6 (Me) (one CH2 not resolved); m/z (CI; NH3) 269
[100%, (M ϩ H)ϩ], 112 (20), 86 (55) and 52 (95) [Found:
(M ϩ H)ϩ, 269.2234. C15H28N2O2 requires M ϩ H, 269.2229].
(2S )-tert-Butyl 2-[2-(1-piperidinyl)ethyl]-1-pyrrolidinecarboxy-
late (S )-51
Representative procedure for evaluating diamines: (R)-2-tri-
methylsilyl-N-tert-butoxycarbonylpyrrolidine (R)-2
Using the procedure described above, protected amino alcohol
(S)-47 (669 mg, 3.11 mmol), methanesulfonyl chloride
(0.27 cm3, 3.42 mmol) and triethylamine (0.66 cm3, 4.66 mmol)
in CH2Cl2 (15 cm3) followed by refluxing with piperidine
(1.25 cm3, 12.5 mmol) in toluene (16 cm3) for 16 h (note: no
DBU added) gave the crude amino carbamate (670 mg, 76%) as
a yellow oil which was sufficiently pure (by TLC and 1H NMR
spectroscopy) for use in the next step.
sec-Butyllithium (2.3 cm3 of a 1.1 M solution in hexane, 2.51
mmol) was added dropwise to a stirred solution of freshly dis-
tilled diamine (ϩ)-24 (488 mg, 2.51 mmol) in Et2O (25 cm3)
at Ϫ78 ЊC under nitrogen. The resulting solution was stirred at
Ϫ78 ЊC for 10 min after which a solution of N-Boc pyrrolidine 1
(331 mg, 1.93 mmol) in Et2O (1.3 cm3) was added dropwise over
10 min via a cannula. After stirring for 5 h at Ϫ78 ЊC, Me3SiCl
(0.37 cm3, 2.90 mmol) was added dropwise and the resulting
solution was allowed to warm to room temperature over 16 h.
Then, 5% aqueous H3PO4 solution (3.2 cm3) was added and the
mixture was stirred for 20 min. The two layers were separated
and the organic layer was washed with 5% aqueous H3PO4 solu-
tion (3 × 3.2 cm3). The combined aqueous layers were extracted
with Et2O (4 × 4.3 cm3) and the combined Et2O extracts were
dried (MgSO4) and evaporated under reduced pressure to give
the crude product. Purification by flash column chromato-
graphy with light petroleum–EtOAc (95 : 5) as eluent gave the
trimethylsilyl adduct (R)-2 (394 mg, 84%, 90% ee by chiral GC)
as a colourless oil, [α]D Ϫ61.7 (c 1.0 in CHCl3) (lit.,7 [α]D ϩ71.8
(c 2.6 in CHCl3) for (S)-5 of 96% ee); GC: Chiraldex G-PN
20 m × 0.25 mm i.d. (γ-cyclodextrin, propionyl derivative in
3-position), 27.0 min [(S)-2], 28.0 min [(R)-2]. Spectroscopic
data identical to that reported in the literature.7
In a separate experiment, a sample of amino carbamate (S)-
51 was purified by flash column chromatography using CH2Cl2–
MeOH (9 : 1) as eluent. Amino alcohol (S)-51: yellow solid,
[α]D Ϫ58.9 (c 1.0 in CHCl3); RF (9 : 1 CH2Cl2–MeOH) 0.35;
νmax(CHCl3)/cmϪ1 1680 (C᎐O); δ (270 MHz; CDCl ) 3.72–3.56
᎐
H
3
(1 H, m), 3.30–3.05 (3 H, m), 3.02–2.65 (5 H, m), 2.13–1.62 (9
H, m), 1.58–1.38 (3 H, m) and 1.26 (9 H, s, CMe3); δC (67.9
MHz; CDCl ) 154.7 (C᎐O), 79.1 (CMe ), 54.7 (CH N), 54.5
᎐
3
3
2
(CHN), 52.8 (CH2N), 46.3 (CH2N), 30.3 (CH2), 28.7 (CH2),
28.0 (CMe3), 23.3 (CH2), 22.3 (CH2) and 21.7 (CH2); m/z (CI;
NH3) 283 [100%, (M ϩ H)ϩ], 183 (27) and 98 (48) [Found: (M ϩ
H)ϩ, 283.2384. C16H30N2O2 requires M ϩ H, 283.2386].
(S )-N-Methyl-[2-(1-piperidinyl)ethyl]piperidine (S )-26
Lithium aluminium hydride (795 mg, 21.0 mmol) was added in
one portion to a stirred solution of amino carbamate (S)-50
(936 mg, 3.5 mmol) in THF (40 cm3) at room temperature
under nitrogen. After stirring for 48 h, Et2O (20 cm3) was added
followed by the portionwise addition (CAUTION) of solid
hydrated sodium sulfate until effervescence ceased. The
resulting suspension was then stirred at room temperature for
3 h and the solids were removed by filtration through Celite.
The filtrate was evaporated under reduced pressure to give the
crude product. Purification by Kugelrohr distillation gave
Acknowledgements
We thank Astra-Zeneca for a fully funded studentship (to
J-P. R. H.), EPSRC and SmithKline Beechams (now Glaxo-
SmithKline) for a CASE studentship (to D. W. P.), BBSRC
for a studentship (to M. J. D.), The Leverhulme Trust for a
fellowship (to J. R. H.), the EU for Erasmus funding (to
T. K. and J. P.) and The Royal Society/NATO for a postdoctoral
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3987

View Article Online
1998, 2151; K. Tomooka, L-F. Wang, N. Komine and T. Nakai,
fellowship award (to V. T.). We are extremely grateful to
Dr D. B. Matthews (of GlaxoSmithKline) for numerous chiral
HPLC and chiral GC analyses. Dr Catherine R. Firkin (formely
Astra-Zeneca) is thanked for her interest in the project. Dr
Ian O’Neil (University of Liverpool) and Dr Iain Coldham
(University of Sheffield) are thanked for useful exchange of
information and discussion.
Tetrahedron Lett., 1999, 40, 6813; S. Nakamura, R. Nakagawa,
Y. Watanabe and T. Toru, J. Am. Chem. Soc., 2000, 122, 11340;
S. Nakamura, A. Furutani and T. Toru, Eur. J. Org. Chem., 2002,
1690; D. M. Hodgson, C. R. Maxwell and I. R. Matthews,
Tetrahedron: Asymmetry, 1999, 10, 1847; D. M. Hodgson, I. D.
Cameron, M. Christlieb, R. Green, G. P. Lee and L. A. Robinson,
J. Chem. Soc., Perkin Trans. 1, 2001, 2161.
34 R. E. Gawley and Q. Zhang, J. Org. Chem., 1995, 60, 5763; R. E.
Gawley, E. Low, Q. Zhang and R. Harris, J. Am. Chem. Soc., 2000,
122, 3344.
35 I. Coldham, S. Dufour, T. F. N. Haxell, S. Howard and G. P. Venall,
Angew. Chem., Int. Ed., 2002, 41, 3887.
36 D. J. Gallagher, S. Wu, N. A. Nikolic and P. Beak, J. Org. Chem.,
1995, 60, 8148.
References
1 S. T. Kerrick and P. Beak, J. Am. Chem. Soc., 1991, 113, 9708.
2 D. Hoppe, F. Hintze and P. Tebben, Angew. Chem., Int. Ed. Engl.,
1990, 29, 1422.
3 For reviews, see: D. Hoppe and T. Hense, Angew. Chem., Int. Ed.
Engl., 1997, 36, 2282; P. Beak, A. Basu, D. J. Gallagher, Y. S. Park
and S. Thayumanavan, Acc. Chem. Soc., 1996, 29, 552; J. Clayden,
Organolithiums: Selectivity for Synthesis, Pergamon, 2002;
T. Schütz, Synlett, 2003, 901 (Spotlight 63: (Ϫ)-Sparteine in
Asymmetric Synthesis).
4 For a selection of the most recent examples, see: M. C. Whisler
and P. Beak, J. Org. Chem., 2003, 68, 1207; D. M. Hodgson,
M. A. H. Stent, B. Stefane and F. X. Wilson, Org. Biomol. Chem.,
2003, 1, 1139; Y. Zhang, S-M. Yeung, H. Wu, D. P. Heller, C. Wu
and W. D. Wulff, Org. Lett., 2003, 5, 1813; S. McN. Sieburth,
H. K. O’Hare, J. Xu, Y. Chen and G. Liu, Org. Lett., 2003, 5, 1859.
5 R. Shintani and G. Fu, Angew. Chem., Int. Ed., 2002, 41, 1057.
6 D. R. Jensen, J. S. Pugsley and M. S. Sigman, J. Am. Chem. Soc.,
2001, 123, 7475; E. M. Ferraira and B. M. Stolz, J. Am. Chem. Soc.,
2001, 123, 7725; D. R. Jensen and M. S. Sigman, Org. Lett., 2003,
5, 63; J. T. Bagdanoff, E. M. Ferraira and B. M. Stolz, Org. Lett.,
2003, 5, 835; J. A. Mueller and M. S. Sigman, J. Am. Chem.
Soc., 2003, 123, 7005; S. K. Mandal, D. R. Jensen, J. S. Pugsley and
M. S. Sigman, J. Org. Chem., 2003, 68, 4600; I. S. Ali and A. Sudulai,
Tetrahedron Lett., 2002, 43, 5435.
37 K. B. Wiberg and W. F. Bailey, Tetrahedron Lett., 2000, 41, 9365.
38 X. Li, L. B. Schenkel and M. C. Kozlowski, Org. Lett., 2000, 2, 875;
Z. Xu and M. C. Kozlowski, J. Org. Chem., 2002, 67, 3072.
39 X. Li, J. Yang and M. C. Kozlowski, Org. Lett., 2001, 3, 1137; X. Li,
B. Hewgley, C. A. Mulrooney, J. Yang and M. C. Kozlowski, J. Org.
Chem., 2003, 68, 5500.
40 For preliminary communication of some of our results, see:
M. J. Dearden, C. R. Firkin, J-P. R. Hermet and P. O’Brien,
J. Am. Chem. Soc., 2002, 124, 11870; J. R. Harrison, P. O’Brien,
D. W. Porter and N. M. Smith, Chem. Commun., 2001, 1202.
41 J. R. Harrison, P. O’Brien, D. W. Porter and N. M. Smith, J. Chem.
Soc., Perkin Trans. 1, 1999, 3623.
42 P. Scheiber and P. Nemes, Liebigs Ann. Chem., 1994, 1033.
43 F. D. King, J. Chem. Soc., Perkin Trans. 1, 1986, 447.
44 A. Barco, S. Benetti, C. De Risi, P. Marchetti, G. P. Pollini and
V. Zanirato, Eur. J. Org. Chem., 2001, 975.
45 R. W. Scott, J. Epperson and C. H. Heathcock, J. Org. Chem., 1998,
63, 5001.
46 A. H. Beckett, R. G. Lingard and A. E. E. Theobald, J. Med. Chem.,
1969, 12, 563.
47 N. J. Leonard, R. W. Fulmer and A. S. Hay, J. Am. Chem. Soc., 1956,
78, 3457.
48 R. Bhide, R. Mortezaei, A. Scilimati and C. J. Sih, Tetrahedron Lett.,
1990, 31, 4827.
49 R. A. Bunce, C. J. Peeples and P. B. Jones, J. Org. Chem., 1992, 57,
1727.
7 P. Beak, S. T. Kerrick, S. Wu and J. Chu, J. Am. Chem. Soc., 1994,
116, 3231.
8 D. J. Gallagher, S. T. Kerrick and P. Beak, J. Am. Chem. Soc., 1992,
114, 5872.
9 D. J. Gallagher and P. Beak, J. Org. Chem., 1995, 60, 7092.
10 K. M. Bertini Gross and P. Beak, J. Am. Chem. Soc., 2001, 123, 315.
11 K. B. Wiberg and W. F. Bailey, Angew. Chem., Int. Ed., 2000, 39,
2127.
50 M. P. Cooke and R. K. Widener, J. Org. Chem., 1989, 52, 1381.
51 R. W. Hoffmann, T. Sander and A. Hense, Liebigs Ann. Chem.,
1993, 771.
12 K. B. Wiberg and W. F. Bailey, J. Am. Chem. Soc., 2001, 123, 8231.
13 N. A. Nikolic and P. Beak, Org. Synth., 1996, 74, 23.
14 J. R. Harrison and P. O’Brien, Synth. Commun., 2001, 31, 1155.
15 M. Majewski, J. Shao, K. Nelson, P. Nowak and N. M. Irvine,
Tetrahedron Lett., 1998, 39, 6787.
16 R. K. Dieter, C. M. Topping, K. R. Chandupatla and K. Lu, J. Am.
Chem. Soc., 2001, 123, 5132.
52 F. Toda, K. Tanaka, H. Ueda and T. Oshima, J. Chem. Soc., Chem.
Commun., 1983, 743.
53 P. J. Wagner, J. H. Sedon and A. Gudmunsdottir, J. Am. Chem. Soc.,
1996, 118, 746.
54 F. Toda, K. Tanaka and H. Ueda, Tetrahedron Lett., 1981, 22, 4669.
55 For example, (Ϫ)-cytisine costs £80.40 for 25 mg from Sigma-
Aldrich Company Ltd.
17 K. M. Bertini Gross, Y. M. Jun and P. Beak, J. Org. Chem., 1997, 62,
7679.
56 D. E. Marrière, J. Rouden, V. Tadino and M-C. Lasne, Org. Lett.,
2000, 2, 1121.
57 N. J. Leonard, in The Alkaloids, ed. R. H. F. Manske and
H. L. Holmes, Academic Press, New York, 1953, vol. 3, pp. 119–199.
58 P. Mascagni, M. Christodoulou, W. A. Gibbons, K. Asres,
J. D. Phillipson, N. Niccolai and S. Mangani, J. Chem. Soc., Perkin
Trans. 2, 1987, 1159.
18 N. Kise, T. Urai and J-I. Yoshida, Tetrahedron: Asymmetry, 1998, 9,
3125.
19 I. Coldham, R. C. B. Copley, T. F. N. Haxell and S. Howard, Org.
Lett., 2001, 3, 3799; I. Coldham, R. C. B. Copley, T. F. N. Haxell and
S. Howard, Org. Biomol. Chem., 2003, 1532.
20 W. F. Bailey, P. Beak, S. T. Kerrick, S. Ma and K. B. Wiberg, J. Am.
Chem. Soc., 2002, 124, 1889.
21 A. Orechoff, M. Rabinowitch and R. Kolowanowa, Ber. Dtsch.
Chem. Ges., 1933, 66, 621.
22 T. Ebner, M. Eichelbaum, P. Fischer and C. O. Meese, Arch. Pharm.
(Weinheim), 1989, 322, 399.
23 N. J. Leonard and R. E. Beyler, J. Am. Chem. Soc., 1950, 72, 1316.
24 B. T. Smith, J. A. Wendt and J. Aubé, Org. Lett., 2002, 4, 2577.
25 M. Schlosser and D. Limat, J. Am. Chem. Soc., 1995, 117, 12342.
26 A. Basu and S. Thayumanavan, Angew. Chem., Int. Ed., 2002, 41,
716.
27 M. Paetow, M. Kotthaus, M. Grehl, R. Fröhlich and D. Hoppe,
Synlett, 1994, 1034.
28 D. Hoppe, M. Paetow and F. Hintze, Angew. Chem., Int. Ed., 1993,
394.
29 T. Kimachi and Y. Takemoto, J. Org. Chem., 2001, 66, 2700.
30 H. Paulsen, C. Graeve and D. Hoppe, Synthesis, 1996, 141; Y. S. Park
and P. Beak, J. Org. Chem., 1997, 62, 1574.
59 J. Rouden, A. Ragot, S. Gouault, D. Cahard, J-C. Plaquevent and
M-C. Lasne, Tetrahedron: Asymmetry, 2002, 13, 1299.
60 A. Padwa, S. M. Sheehan and C. S. Straub, J. Org. Chem., 1999, 64,
8648; J. M. McIntosh, L. Z. Pillon, S. O. Acquaah, J. R. Green and
G. S. White, Can. J. Chem., 1983, 61, 2016.
61 H. Kubo, S. Ohmiya and I. Murakoshi, Can. J. Chem., 1994, 72, 214.
62 J. R. Harrison and P. O’Brien, Tetrahedron Lett., 2000, 41, 6167.
63 M. S. Toy and C. C. Price, J. Am. Chem. Soc., 1960, 82, 2613.
64 C. Morley, D. W. Knight and A. C. Share, J. Chem. Soc., Perkin
Trans. 1, 1994, 2903.
65 I. A. O’Neil, C. L. Murray, A. J. Potter and S. B. Kalinddjian,
Tetrahedron Lett., 1997, 38, 3609.
66 S. Abele, K. Vögtli and D. Seebach, Helv. Chim. Acta, 1999, 82,
1539.
67 I. A. O’Neil, University of Liverpool, personal communication.
68 W. Carruthers and R. C. Moses, J. Chem. Soc., Perkin Trans. 1, 1988,
2251; S. Aoyagi, T. C. Wang and C. Kibyashi, J. Am. Chem. Soc.,
1993, 115, 11393.
31 R. Wilhelm, I. K. Sebhat, A. J. P. White, D. J. Williams and
D. A. Widdowson, Tetrahedron: Asymmetry, 2000, 11, 5003.
32 Y-L. Tan, A. J. P. White, D. A. Widdowson, R. Wilhelm and
D. J. Williams, J. Chem. Soc., Perkin Trans. 1, 2001, 3269.
33 D. M. Hodgson, G. P. Lee, R. E. Marriott, A. J. Thompson,
R. Wisedale and J. Witherington, J. Chem. Soc, Perkin Trans. 1,
69 S. K. Hendrie and J. Leonard, Tetrahedron, 1987, 43, 3289.
70 B. Danieli, G. Lesma, D. Passarella, P. Piacenti, A. Sacchetti,
A. Silvani and A. Virdis, Tetrahedron Lett., 2002, 43, 7155–7158.
71 D. M. Hodgson and S. L. M. Norsikian, Org. Lett., 2001, 3, 461.
72 A. F. Burchat, J. M. Chong and N. Nielson, J. Organomet. Chem.,
1997, 542, 281.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 3 9 7 7 – 3 9 8 8
3988