Tandem radical cyclization reactions, initiated at nitrogen, as an approach to the CDE-tricylic cores of certain post-secodine alkaloids

Article abstract of DOI:10.3987/COM-05-10575

-The nitrogen-radical precursors (10-15) have been prepared and subjected to reaction conditions expected to promote tandem radical cyclization sequences leading to the CDE-tricyclic frameworks associated with alkaloids such as vindoline (1) and ibophylli

Full text of DOI:10.3987/COM-05-10575

HETEROCYCLES, Vol. 68, No. 1, 2006
71
HETEROCYCLES, Vol. 68, No. 1, 2006, PP. 71 – 92. © The Japan Institute of Heterocyclic Chemistry
Received, 26thSeptember, 2005, Accepted, 11th November, 2005, Published online, 15th November, 2005. COM-05-10575
TANDEM RADICAL CYCLIZATION REACTIONS, INITIATED AT
NITROGEN, AS AN APPROACH TO THE CDE-TRICYLIC CORES OF
CERTAIN POST-SECODINE ALKALOIDS
Martin G. Banwell* and David W. Lupton
Research School of Chemistry, The Australian National University, Canberra,
ACT 0200, Australia. E-mail: mgb@rsc.anu.edu.au
Abstract – The nitrogen-radical precursors (10–15) have been prepared and
subjected to reaction conditions expected to promote tandem radical cyclization
sequences leading to the CDE-tricyclic frameworks associated with alkaloids
such as vindoline (1) and ibophyllidine (2). In the event, each of precursors (10,
11, 12 and 13) participated in the desired processes and thus providing products,
(28, 29, 30 and 31) respectively, embodying ring systems related to ibophyllidine
(2). In contrast, the higher homologue (14) of PTOC-carbamate (12) decomposed
upon exposure to radical chain initiation conditions while the N-chloro-analogue
(15) simply underwent reductive dechlorination to give compound (27). As such
the title processes appear unlikely to offer a useful approach to the CDE-tricyclic
ring system associated with vindoline-type alkaloids.
INTRODUCTION
The monoterpenoid indole alkaloids (+)-vindoline (1) and (+)-ibophyllidine (2) are members of the
so-called “post-secodine” class.1 The former natural product has been isolated, as a major constituent,
from various plants including Catharanthus (Vinca) rosea, Catharanthus ovalis and Melodinus
fusiformis.1b,2 Significantly, compound (1) embodies one half of and is a biosynthetic precursor to the
clinically important indole-indoline alkaloids vinblastine and vincristine.3 Unlike the latter alkaloids,
however, vindoline does not display significant anti-mitotic activity nor is it a particularly cytotoxic
entity.4 Ibophyllidine (2) has been obtained from, inter alia, the plants Tabernaemontana albiflora and
Tabernanthe iboga.5 Extracts of the roots of the latter plant are used in West Africa as a ceremonial
hallucinogen or, in smaller doses, as a stimulant. Biogenetically speaking, compound (2) is probably

72
HETEROCYCLES, Vol. 68, No. 1, 2006
derived from the hydroxylated D-ring homologue pandoline.6 While they differ in the nature and size of
the D-ring, alkaloids (1) and (2) have many structural features in common – most conspicuously the
overlap of their ABCE-substructures. These similarities are, clearly, a consequence of their closely related
biogenetic origins.
Vindoline has been the subject of numerous preparative studies, including more than ten successful total
or formal total syntheses7 and much of the more recent work in the area has been reviewed.1a The most
recently reported (and formal) synthesis was described by Murphy and co-workers7a who exploited a
tandem radical cyclization process involving, as the second step, addition of a C-centered radical onto a
tethered azide. Subsequent loss of dinitrogen then hydrogen atom abstraction by the resulting N-centered
radical completed the assembly of a tetracyclic ring system corresponding to the ABCE-core associated
with vindoline (1). This sequence allowed for access to an advanced intermediate associated with Büchi’s
original synthesis of the racemic modification of the alkaloid.7k Much less work has been focused on the
development of syntheses of ibophyllidine with Kuehne and co-workers representing the only group to
have, thus far, achieved total syntheses of the (±)- and then the (+)-form of this alkaloid.6a,8 The work
described herein was undertaken as part of a more general program9 to develop syntheses of a range of
monoterpenoid indole alkaloids including aspidospermidine (3) and, ultimately, vinblastine. A particular
focus of the efforts detailed below was to establish protocols wherein an intact C-ring precursor to
compound (1 and/or 2) could be used as the starting point in the synthesis of these natural products. This
focus was prompted by having access to large quantities of an enantiomerically pure metabolite
embodying functionalities closely related to those observed within the C-ring of vindoline.10
N
N
N
D
D
E
E
A
A
MeO
C
C
B
B
OAc
OH
CO2Me
N
N
N
H
H
H
CO2Me
Me
H
1
2
3
The tandem N-radical initiated cyclization studies of Zard, as deployed in his recently reported synthesis
of (±)-13-deoxyserratine,11 provided the inspiration for the work detailed herein. In particular, it seemed
reasonable to expect that an N-centered radical of the general type (4) would engage in a 5-exo-trig
cyclization reaction to give the perhydroindole-based radical (5) (see Figure 1). The latter species could
then engage in a second cyclization reaction leading to a six-membered ring (via either a 6-endo-trig or
6-endo-dig cyclization reaction depending on the nature of the tethered C–C multiple bond) or a
five-membered ring (via a 5-exo-trig or 5-exo-dig cyclization process). The first mode of cyclization

HETEROCYCLES, Vol. 68, No. 1, 2006
73
(Path a) would be expected to lead, via intermediate radical (6), to the tricyclic species (7) embodying the
CDE-ring substructure of vindoline and related alkaloids of this class. Alternately, if Path b were
followed then radical (8) so formed would be expected to deliver the tricyclic compound (9) incorporating
the CDE-ring system associated with ibophyllidine (2). The stereochemical outcomes of the initial steps
of both sequences were expected to be analogous to those observed by Cossy et al.12 during the course of
their development of an N-centered radical cyclization route to the cis-ring fused perhydroindole
substructure associated with the alkaloidal framework γ-lycorane and corresponding to the CE-ring
substructure within compounds (7 and 9). With such ring-fusion established in the first steps of the
reaction sequences the illustrated stereochemical relationship between the C- and D-rings, as established
in the second cyclization event, necessarily follows. Of course, if the alkyne-containing cyclization
precursor (4) were successfully employed in such processes then the unsaturation introduced into/onto the
D-rings of each of compounds (7 and 9) would lead to functionality at these sites relevant to the synthesis
of targets (1 and 2). The radical cyclization sequences proposed here are essentially the inverse of those
employed by Murphy and his co-workers in developing their recent formal total synthesis of vindoline.7a
N
N
6-endo-
trig/dig
H-atom
donor
D
E
R
R
C
Path a
6
7
N
N
R
R
5-exo-trig
4
5
N
N
H-atom
donor
D
R = H or alkyl
E
Path b
R
R
C
5-exo-
trig/dig
8
9
Figure 1.
While both aminyl and amidyl radicals can be used in N-centered radical cyclization reactions it was
decided to focus attention on the former species as they have been examined more extensively and would,
hopefully, provide direct access to the target tricyclic amines (as opposed to the corresponding lactams).13
Many routes to aminyl radicals have been established since the pioneering work of Stella and Surzur13p,r
but the use of N-chloroamines and N-hydroxypyridine-2-thione carbamates (PTOC carbamates), as
developed by Tokuda13a,e,h and Newcomb13f,k,l,m respectively, seemed especially attractive protocols. So,
for example, it has been demonstrated, by Tokuda, that when N-chloroamines are exposed, in refluxing

74
HETEROCYCLES, Vol. 68, No. 1, 2006
benzene, to tri-n-butyltin hydride and the radical initiator AIBN then a neutral N-radical is formed and
this might be expected to cyclize onto the pendant cyclohexenyl double bond. PTOC-carbamates behave
in a similar manner to Barton esters and can be photolyzed, in the presence of either a Lewis or Brønsted
acid, to generate cationic N-radicals that can also participate in cyclization reactions of the desired type.
Interestingly, when the acidic additive (e.g. malonic acid) is omitted from these types of reactions then the
cyclization reaction does not proceed and the major product is the amine arising from reductive cleavage
of the PTOC moiety.13f,k,l,m
On the basis of the foregoing discussion, the N-radical precursors sought for the purposes of undertaking
the present study were the 2-(cyclohex-2-enyl)ethylamine systems (10–15). This series of compound
varies in, (i), the nature of the second carbon-based (and unsaturated) substituent at nitrogen (propargyl,
allyl or homopropargyl), (ii) the precursor to the N-centered radical (N-chloroamine or PTOC carbamate)
and, (iii) the presence or absence of an ethyl group on the cyclohexenyl double bond. This last feature
allows for an investigation of the capacity of the proposed tandem radical cyclization sequence to
construct tertiary and quaternary carbon centers at the junction between the C- and D-rings of target (1).
The following section details the protocols established for the synthesis of the target substrates (10–15)
while the one thereafter describes studies associated with efforts to engage these materials in the proposed
cyclization sequences.
N
N
S
O
S
O
O
O
Cl
Cl
Cl
1'
Cl
1
1
1'
N
N
N
N
N
N
10
11
12
13
14
15
RESULTS AND DISCUSSION
Synthesis of Substrates for Cyclization Studies
The synthesis of the first of the target substrates (10) was accomplished using the reaction sequence
defined in Scheme 1. Thus, following the five-step protocol reported by Cossy et al.12 and involving a
Pd[0]-catalyzed allylic substitution reaction as the key process, the commercially available
2-cyclohexenone (16) was converted into the previously reported aldehyde (17) (60% from 16).12,14 The
latter compound was then condensed with propargylamine in the presence of 4 Å molecular sieves and

HETEROCYCLES, Vol. 68, No. 1, 2006
75
the resulting imine immediately reduced, using NaBH4 in methanol, to the corresponding secondary
amine 18 which was obtained in 94% yield over the two steps involved. Reaction of compound (18) with
N-chlorosuccinimide in CH2Cl2 then afforded the target cyclization substrate (10) in quantitative yield.
1
13
Satisfactory H and C NMR spectral data could be obtained on this chromatographically stable material
and an accurate MS measurement consistent with the expected molecular formula was recorded on the
[M–H·]+ ion observed in the electron impact MS spectrum. The significant downfield shift (Δδ ca. 0.4) of
the resonances due to the C–1 and C–1' protons observed on converting amine (18) into its
N-chloro-derivative was taken as a clear indication that the desired compound (10) had indeed been
formed.
H
1'
1
N
CHO
O
i
ii,iii
iv
10
17
16
18
Scheme 1. Reagents and conditions: (i) see reference 12; (ii) HC≡CCH2NH2 (5 mol equiv.), 4 Å
molecular sieves, THF, 18°C, 16 h; (iii) NaBH4, (1 mol equiv.), methanol, 0°C, 0.5 h; (iv)
N-chlorosuccinimide (1.2 mol equiv.), CH2Cl2, 0°C, 1 h.
The ethylated equivalent of substrate (10), namely compound (11), was obtained from commercially
available 3-ethoxycyclohexenone (19) by the route indicated in Scheme 2 and which represents, in its
opening stages, a straightforward adaptation of the reaction sequence used by Cossy et al.12 in their
synthesis of aldehyde (17). Thus, reaction of compound (19) with ethylmagnesium bromide and
subjection of the resulting tertiary alcohol to acidic workup then gave the previously reported15
3-ethylcyclohexenone (20) in 89% yield. Sodium borohydride-mediated reduction of the latter compound
afforded the corresponding allylic alcohol which was immediately acetylated, under standard conditions,
to give the previously reported16 allylic acetate (21) in quantitative yield over the two steps just described.
Treatment of compound (21) with Pd2dba3/1,2-dppe and the sodium salt of dimethyl malonate in refluxing
THF then afforded the allylic substitution product (22) in a completely regioselective manner and 90%
yield (at 74% conversion). Following a procedure reported by Wenkert,17 this diester was treated with
lithium iodide in DMSO at 180°C and by such means one of the carbomethoxy groups was removed and
thus affording the monoester (23) in 80% yield. Reduction of this last compound with DIBAL-H in
CH2Cl2 at –78°C afforded the corresponding aldehyde (24) (75%) and this was then subjected to a
reductive amination sequence using propargylamine. The secondary amine (25) (75%) so-formed, and for
which satisfactory 1H and 13C NMR spectral data were obtained, was then reacted with

76
HETEROCYCLES, Vol. 68, No. 1, 2006
N-chlorosuccinimde, in same manner as detailed above, to give the N-chloroamine (11) (quant.) which
could only be partially characterized because of its rather unstable nature. The equivalent
PTOC-derivative (12) was prepared (Scheme 2) by reacting the same amine with
1
1-oxa-2-oxo-3-thiaindolizinium chloride18 according to method of Newcomb et al.13f In the 300 MHz H
NMR spectrum of this material the C–1' protons resonated as two distinct singlets (at δ 4.43 and 4.21)
while the C–1 protons appeared as two distinct triplets (at δ 3.72 and 3.53). Such “splittings” suggest this
compound exists in two distinct forms arising from restricted rotation about the N to C=O bond
associated with the newly installed carbamate residue.
CO2Me
EtO
O
AcO
O
i
ii,iii
iv
MeO2C
20
19
21
22
v
H
viii
ix
11
12
CHO
CO2Me
N
vii
vi
25
24
23
x
H
N
viii
13
26
Scheme 2. Reagents and conditions: (i) EtMgBr (1.66 mol equiv.), THF, 0°C, 3 h; (ii) NaBH4 (1.2 mol
equiv.), methanol, 0 to 18°C, 16 h; (iii) Ac2O, pyridine, DMAP (cat.), 0 to 18°C, 10.5 h; (iv)
(MeO2C)2CHNa (1.7 mol equiv.), Pd[0] (cat.), THF, 68°C, 24 h; (v) LiI•3H2O (1 mol equiv.), DMSO,
180°C, 1.5 h; (vi) DIBAL-H (1.1 mol equiv.), CH2Cl2,/hexane, –78°C, 1.5 h; (vii) HC≡CCH2NH2 (5 mol
equiv.), 4 Å molecular sieves, THF, 18°C, 16 h, then NaBH4, MeOH, 0°C; (viii) N-chlorosuccinimide
(1.2 mol equiv.), CH2Cl2, –30 to 18°C, 1 h; (ix) 1-oxa-2-oxo-3-thiaindolizinium chloride (1.2 mol equiv.),
Et3N (1.1 mol equiv.), C6H6, 18°C, 2 h; (x) H2C=CCH2NH2 (5 mol equiv.), 4 Å molecular sieves, THF,
18°C, 4 h, then NaBH4, MeOH, 0°C.

HETEROCYCLES, Vol. 68, No. 1, 2006
77
The preparation of the cyclization substrate (13) was achieved (Scheme 2) by subjecting the
above-mentioned aldehyde (24) to reaction with allylamine in the presence of 4Å molecular sieves and
the resulting imine immediately reduced with NaBH4. In this manner compound (26) was obtained in
75% yield. Reaction of the last species with N-chlorosuccinimide in the same manner as described above
then afforded the targeted N-chloroamine (13) in 97% yield. The unstable nature of compound (13)
precluded the acquisition of MS spectral data but the derived NMR spectral data (see EXPERIMENTAL)
were in full accord with the assigned structure and similar to those obtained on congener (11).
The final substrates (14 and 15), sought in connection with the present study were prepared by the route
defined in Scheme 3. The approach involved was slightly different from that employed in obtaining
congeners (11 and 12) because, despite many attempts, we were unable to engage homopropargylamine
in a Schiff-base condensation with aldehyde (24). However, the ylide derived from homopropargyl azide
and triphenyphosphine was readily obtained and this reacted with aldehyde (24) to form the necessary
imine.19 The so-obtained and crude samples of this material were treated with sodium cyanoborohydride
and upon work up then chromatography the, by now long sought after, amine (27) was finally obtained,
albeit in only 33% yield. This material was then converted, in high yields, into the corresponding PTOC-
and N-chloro-derivatives (14 and 15) respectively, using the protocols defined above for the generation of
their lower homologues. While the former product was so unstable that no useful spectroscopic analysis
1
could be carried out, the 300 MHz H NMR spectrum of compound (15) was acquired and proved fully
consistent with the assigned structure.
ii
14
H
i
N
24
iii
15
27
Scheme 3. Reagents and conditions: (i) H2C≡CCH2CH2N3 (1 mol equiv.), PPh3 (1 mol equiv.), THF,
68°C, 16 h then NaCNBH3 (1.5 mol equiv.); (ii) 1-oxa-2-oxo-3-thiaindolizinium chloride (1.2 mol equiv.),
Et3N (1 mol equiv.), C6H6, 18°C, 2 h; (iii) N-chlorosuccinimide (1.1 mol equiv.), CH2Cl2, –30 to 18°C,
1 h.
Cyclization Studies
The results of subjecting substrates (10–15) to the foreshadowed cyclization conditions are shown in
Table 1 as well as Schemes 4–6. Thus, treatment of the first of these compounds, N-chloroamine (10), at
40 mmolar concentration to reaction with n-Bu3SnH in refluxing toluene and in the presence of AIBN
(Entry 1) afforded the dechlorinated compound (18) (41%) as the major product of reaction but the

78
HETEROCYCLES, Vol. 68, No. 1, 2006
anticipated and chromatographically separable tandem radical cyclization product (28) (Scheme 4) was
also obtained in 34% yield. Despite the less than complete mass balance encountered in this reaction, no
evidence for the formation of other cyclization products was obtained. The samples of compound (18)
1
13
generated in this manner were identical, as judged by H and C NMR spectral analyses, to those
obtained earlier. Spectral data derived from the latter product (28) were in full accord with the assigned
structure. In particular, the electron-impact MS spectrum displayed a molecular ion at m/z 163 and an
accurate MS measurement on this species confirmed that it was of the expected composition, viz. C11H17N.
1
While the IR spectrum was rather non-descript, the 300 MHz H NMR spectrum revealed, as the most
significant feature, two slightly broadened one-proton singlets at δ 4.81 and 4.78 which are assigned to
the hydrogens attached to the terminal and exocyclic olefin associated with tricyclic compound (28). The
75 MHz 13C NMR spectrum was even more informative, displaying the expected eleven signals including
two due to sp2-hydridized carbons associated with an exocyclic alkene and three due to sp3-hydridized
carbons attached to nitrogen. Such data compare favorably with those reported for a structurally related
compound prepared by Kuehne and co-workers.20 These spectral features clearly demonstrate that a
tandem radical cyclization process has taken place with the second of the steps involved occurring in a
5-exo-trig manner (i.e. following Path b as shown in Figure 1). While rigorous proof of stereochemistry
for compound (28) has not been obtained, the illustrated structure is the only one that can reasonably
Table 1: Outcomes of the Subjection of Cyclization Substrates (10–15) to Conditions for Generating
N-Centered Radicalsa
Entry
Substrate (mmol conc.)
Reaction Conditions
Product(s) (yield %)
1
2
10 (40)
10 (100)
10 (20)
10 (10)
11 (60)
12 (50)
12 (50)
13 (40)
13 (10)
14 (25)
14 (25)
15 (27.5)
n-Bu3SnH, AIBN, toluene, reflux
n-Bu3SnH, AIBN, benzene, reflux
n-Bu3SnH, AIBN, benzene, reflux
n-Bu3SnH, AIBN, benzene, reflux
n-Bu3SnH, AIBN, toluene, reflux
BF3•Et2O, hν, acetonitrile
18 (41), 28 (34)
18 (63), 28 (19)
18 (50), 28 (27)
18 (0), 28 (74)
29 (68)
3
4
5
6
30 (71)
Malonic acid, hν, acetonitrile
n-Bu3SnH, AIBN, toluene, reflux
n-Bu3SnH, AIBN, benzene, reflux
BF3•Et2O, hν, acetonitrile
7
30 (65)
8
26 (35), 31 (44)
26 (1), 31 (62)
Decomposition
Decomposition
27 (~50)
9
10
11
12
Malonic acid, hν, acetonitrile
n-Bu3SnH, AIBN, toluene, reflux
a
Full details provided in the EXPERIMENTAL.

HETEROCYCLES, Vol. 68, No. 1, 2006
79
accommodate the fusion of two five-membered rings across contiguous positions of a six-membered ring
system. Unsurprisingly, the concentration of substrate (10) used in effecting the reductive cyclization
process has a significant impact on the ratio of products (18 and 28) observed. For example, when 100
mmolar concentrations of substrate were used (Entry 2, Table 1) along with benzene as the reaction
solvent then the reductive dechlorination product (18) was obtained in 63% yield and the yield of target
(28) dropped to 19%. While lowering the concentration of substrate to 20 mmol (Entry 3, Table 1) had
little useful effect, when the reaction was run at the 10 mmolar level (Entry 4, Table 1) the tricyclic
species (28) became the exclusive product of reaction and was obtained in 74% yield.
N
i
10
18
+
28
Scheme 4. Reagents and conditions: (i) n-Bu3SnH (1 mol equiv.), AIBN (5 mol %), C6H6, 80°C, 1.5 h.
Cyclization of the ethyl-substituted counterpart to compound (10), viz. the N-chloroamine (11), at 60
mmolar concentrations in refluxing toluene with n-Bu3SnH and using AIBN as the radical initiator (Entry
5, Table 1) led to the tricyclic amine (29) (68%) as the only observed product of reaction (Scheme 5).
This outcome, when compared to that observed for the equivalent process involving substrate (10) (Entry
1, Table 1), suggests that the presence of the ethyl group on the double-bond being attacked by the
initially formed N-radical facilitates the first of the two cyclization steps involved in these types of
tandem radical cyclization processes. As before, the spectral data obtained on compound (29) were in full
accord with the assigned structure. Interestingly, when a 50 mmolar solution of the PTOC-carbamate
analogue of N-chloroamine (11), namely compound (12), in acetonitrile was subjected to irradiation with
a 250 W tungsten lamp in the presence of equimolar amounts of BF3•Et2O or malonic acid (Entries 6 and
7), as specified in work reported by Newcomb and co-workers, then the thioenol ether (30) (65–71%) was
1
obtained (Scheme 5) as a single geometric isomer and as the only isolable reaction product. In the H
NMR spectrum of this product resonances due to the thiopyridine reside were observed between δ 8.42
and 6.99 while the olefinic proton gave rise to a one-proton triplet (J 2.1 Hz) at δ 6.25. The diastereotopic
and mutually coupled (J 16.2 Hz) methylene protons immediately adjacent to the double bond resonated
at δ 3.94 and 3.40. In the 70 eV electron-impact MS spectrum only a weak molecular ion was observed
but an accurate MS measurement could still be recorded on this species. The base peak, at m/z 190,
corresponds to that fragment arising from the loss of the thiopyridine radical from the molecular ion.

80
HETEROCYCLES, Vol. 68, No. 1, 2006
N
S
N
N
i
ii
11
12
29
30
Scheme 5. Reagents and conditions: (i) n-Bu3SnH (1 mol equiv.), AIBN (20 mol %), C6H6, 80°C, 5 h; (ii)
malonic acid (3 mol equiv.) or BF3•Et2O (3 mol equiv.), acetonitrile, irradiation, 0.5 h, ca. 18°C.
Subjection of the N-chloroamine (13) to reaction with n-Bu3SnH and AIBN in refluxing toluene at 40
mmolar concentrations (Entry 8, Table 1) led to a chromatographically separable mixture of the
dechlorinated amine (26) (35%) and the reductive cyclization product (31) (44%) (Scheme 6). A related
experiment conducted in benzene at 10 mmolar concentrations (Entry 9, Table 1) afforded the latter
compound as the essentially exclusive product of reaction and in 68% yield. In each instance compound
(31) was obtained as a single diastereoisomer and the illustrated structure in which both the ethyl and
methyl substituents reside on the exo-face of the tricyclic framework follows from mechanistic
considerations. In particular, at the transition state associated with the second of the two cyclization steps
involved, and the one in which the stereochemistry of the ring substituents is determined, we assume that
as the favored all-cis-fused tricyclic ring system is developing the ethyl and incipient methyl substituents
are both in the thermodynamically favored exo-orientations. However, rigorous independent proof of the
illustrated stereochemistry within product (31) has not been obtained.
N
i
+
26
13
31
Scheme 6. Reagents and conditions: (i) n-Bu3SnH (1.2 mol equiv.), AIBN (20 mol %), C6H6, 80°C, 6 h.
In an effort to establish routes to the CDE-tricyclic core associated with vindoline (1), the
homopropargyl-containing PTOC-carbamate (14) was subject to Newcomb’s radical cyclization
conditions as successfully deployed in the conversion 12 → 30. In the event, however, when the relevant
reactions were carried out (Entries 10 and 11, Table 1) only decomposition of the substrate was observed.
When the N-chloro-analogue of substrate (14), namely compound (15), was subjected to reaction with
n-Bu3SnH and AIBN in refluxing toluene at ca. 30 mmolar concentrations then only the dechlorinated
amine (27) was observed (50% yield). This disappointing outcome contrasts dramatically with the
behavior of the lower homologue of compound (15), viz. N-chloroamine (11), which cyclized in an

HETEROCYCLES, Vol. 68, No. 1, 2006
81
efficient manner to give product (29) embodying the CDE-tricyclic core associated with alkaloids such as
ibophyllidine (2). The divergent behaviors of compounds (11 and 15) probably arises because the
cyclization processes each engages in are reversible with the second one being rather slow (in the forward
direction) in the case of the higher homologue.13m,21 As a consequence, the equilibrium concentration of
nitrogen radical derived from compound (15) will be much higher than those of the corresponding
carbon-centered isomers with the result that only the direct reduction product (27) is observed. The
situation is essentially reversed in the case of those radical cyclization processes involving compounds
(10, 11 and 12). As a result the tricyclic products (28, 29 and 30) become the major products of the
reactions of these substrates.
CONCLUSIONS
The successful radical cyclization reactions of compounds (10–13) clearly indicate that the title processes
offer a useful means of obtaining compounds (28, 29, 30 and 31), respectively, embodying the
CDE-tricyclic core associated with alkaloids such as ibophyllidine (2). Indeed, the efficiency of such
conversions suggests that the tandem radical cyclization reactions involved could provide an effective
basis for establishing concise routes to this and related alkaloids, the biological profiles of which remain
ill defined. Work directed towards such ends are now underway in these laboratories and will be reported
upon in due course.
In contrast to the success associated with studies involving the cyclization of substrates (10–13), the lack
of useful outcomes encountered during analogous studies on their higher homologues (14 and 15)
highlights the likely inability of tandem radical cyclization reactions of the type defined in Figure 1 to
deliver the CDE-tricyclic core associated with alkaloids such as vindoline. The ineffectiveness of this
approach is probably the consequence of the reversible nature of the cyclization processes involved and
the unfavorable kinetics associated with the relevant radical cyclization and/or chain termination events.
EXPERIMENTAL
General Experimental Procedures. Proton (1H) and carbon (13C) NMR spectra were recorded on a
Varian Gemini 300 NMR spectrometer, operating at 300 MHz (for proton) and 75 MHz (for carbon).
Unless otherwise specified, spectra were acquired at 20°C in deuterochloroform (CDCl3) that had been
filtered through basic alumina immediately prior to use. Chemical shifts are recorded as δ values in parts
per million (ppm). IR spectra (νmax) were recorded on a Perkin–Elmer 1800 Series FTIR Spectrometer and
samples were analyzed as thin films on KBr plates. Low resolution MS spectra were recorded on a
Micromass–Waters LC-ZMD single quadrupole liquid chromatograph-MS or VG Quattro II triple
quadrupole MS instrument using electron impact techniques. High resolution MS spectra were recorded

82
HETEROCYCLES, Vol. 68, No. 1, 2006
on an AUTOSPEC spectrometer. Dichloromethane (CH2Cl2) was distilled from calcium hydride and THF
was distilled, under nitrogen, from sodium benzophenone ketyl. Where necessary, reactions were
performed under a nitrogen atmosphere.
Section A: Synthesis of Substrates for Radical Cyclization Studies
(2-Cyclohex-2-enylethyl)prop-2-ynylamine (18).
A magnetically stirred solution of aldehyde (17)12,14 (905 mg, 7.3 mmol) in THF (25 mL) was treated with
propargylamine (2.00 g, 36.5 mmol) then activated 4 Å molecular sieves (650 mg). The ensuing mixture
was stirred at 18°C for 16 h then filtered through a pad of Celite™ contained in a sintered-glass funnel.
The filtrate was concentrated under reduced pressure and the pale-yellow oil so obtained was dissolved in
methanol (15 mL) and the resulting and magnetically stirred solution cooled to 0°C then treated, in one
portion, with NaBH4 (280 mg, 7.3 mmol). After 0.5 h the reaction mixture was poured into NaOH (100
mL of a 2 M aqueous solution) then extracted with ethyl acetate (3 × 100 mL). The combined organic
extracts were then washed with brine (1 × 100 mL) before being dried (MgSO4), filtered and concentrated
under reduced pressure to yield the title amine (18) (1.10 g, 94%) as a clear, colorless oil [Found:
1
[(M–H·)+, 162.1283. C11H17N requires [(M–H·)+, 162.1283]. H NMR (300 MHz) δ 5.66–5.55 (complex
m, 1H), 5.48 (dm, J 9.9 Hz, 1H), 3.34 (d, J 2.4 Hz, 2H), 2.66 (t, J 7.2 Hz, 2H), 2.15 (t, J 2.4 Hz, 1H), 2.07
(m, 1H), 1.88 (m, 2H), 1.78–1.10 (complex m, 7H); 13C NMR (75 MHz) δ 131.4, 126.9, 82.1, 71.1, 46.0,
37.9, 36.1, 32.9, 28.8, 25.1, 21.2; IR (neat) vmax 3297, 3021, 2927, 2847, 1447, 1432, 1327, 1114 cm–1;
MS (EI) m/z 162 [(M–H·)+, 20%], 161 (10), 108 (50), 93 (40), 82 (100), 68 (95).
N-Chloro-(2-cyclohex-2-enylethyl)prop-2-ynylamine (10).
A magnetically stirred solution of propargylamine (18) (1.10 g, 6.75 mmol) in CH2Cl2 (40 mL)
maintained at 0°C was treated, in one portion, with N-chlorosuccinimide (1.07 g, 8.1 mmol). After 1 h the
entire reaction mixture was added to the top of a pad of flash chromatographic-grade silica gel contained
in a sintered-glass funnel and the pad was then eluted with 3:7 v/v ethyl acetate/hexane. Concentration of
the appropriate fractions (Rf 0.35 in 1:9 v/v ethyl acetate/hexane) under reduced pressure then afforded the
title N-chloroamine (10) (1.32 g, 100%) as a clear, colorless oil [Found: (M–H·)+, 196.0891. C11H1635ClN
requires (M–H·)+, 196.0893]. 1H NMR (300 MHz) δ 5.65 (m, 1H), 5.51 (dm, J 9.9 Hz, 1H), 3.79 (d, J 2.4
Hz, 2H), 3.02 (t, J 7.2 Hz, 2H), 2.39 (t, J 2.4 Hz, 1H), 2.15 (m, 1H), 1.92 (m, 2H), 1.79–1.38 (complex m,
5H), 1.20 (m, 1H); 13C NMR (75 MHz) δ 131.0, 127.3, 77.4, 74.7, 59.6, 52.3, 34.1, 32.6, 28.8, 25.1, 21.1;
IR (neat) vmax 3297, 3014, 2913, 2847, 1645, 1436, 1320, 1255, 1165, 1082 cm–1; MS (EI) m/z 198 and
196 [(M–H·)+, both 10%], 162 (27), 160 (17), 120 (70), 68 (100).
3-Ethylcyclohex-2-en-1-one (20).
Ethylmagnesium bromide (100 mL of a 3 M solution in ether, 0.30 mol, ex. Aldrich Chemical Company)

HETEROCYCLES, Vol. 68, No. 1, 2006
83
was added dropwise to a magnetically stirred solution of 3-ethoxycyclohex-2-en-1-one (25.0 g, 0.18 mol,
ex. Aldrich Chemical Company) in THF (300 mL) maintained at 0°C. After 3 h the by now burnt-orange
colored solution was poured into ice-cold HCl (500 mL of a 10% w/v aqueous solution) and the resulting
yellow suspension stirred for 16 h at 18°C. The ensuing mixture was then extracted with ether (3 × 900
mL), washed with water (1 × 200 mL), NaOH (2 × 200 mL of a 5 % w/v aqueous solution), water (1 ×
200 mL) and brine (1 × 200 mL) then dried (MgSO4), filtered and concentrated under reduced pressure to
give compound (20)15 (19.7 g, 89%) as a clear, colorless oil, Rf 0.5 (3:7 v/v ethyl acetate/hexane) (Found:
M+•, 124.0893. Calcd for C8H12O M+•, 124.0888). 1H NMR (300 MHz) δ 5.75 (s, 1H), 2.28–2.09 (complex
13
m, 6H), 1.88 (m, 2H), 0.99 (t, J 7.5 Hz, 3H); C NMR (75 MHz) δ 199.6 (C), 167.6 (C), 124.2 (CH),
37.1 (CH2), 30.6 (CH2), 29.4 (CH2) 22.4 (CH2), 10.9 (CH3); IR (neat) vmax 2969, 2938, 2879, 1670, 1625,
1458, 1429, 1374, 1348, 1283, 1255, 1192, 1137, 887 cm–1; MS (EI) m/z 124 (M+•, 60%), 105 (15), 96
(100), 91 (50), 81 (47), 79 (44), 77 (42), 65 (31).
3-Ethylcyclohex-2-enyl Acetate (21).
A magnetically stirred solution of enone (20) (19.0 g, 153 mmol) in methanol (250 mL) maintained at
0°C was treated, in one portion, with sodium borohydride (6.98 g, 184 mmol). After the initially rapid
reaction had subsided, the reaction mixture was allowed to warm to 18°C and after a further 16 h it was
treated with water (250 mL) then ether (500 mL). The separated aqueous phase was extracted with ether
(3 × 500 mL) and the combined organic phases were then washed with brine (1 × 200 mL) before being
dried (MgSO4), filtered and concentrated under reduced pressure to yield the expected allylic alcohol
(18.0 g, 96%). This was immediately converted into the corresponding acetate. Thus, a magnetically
stirred solution of the allylic alcohol (18.0 g, 147 mmol) and pyridine (20 mL) in acetic anhydride (30
mL) maintained at 0°C was treated with DMAP (300 mg). When the initially rapid reaction had subsided,
the reaction mixture was allowed to warm to 18°C. After a further 10 h excess acetic anhydride was
destroyed by adding methanol (100 mL) to the ice-cooled reaction mixture. After a further 4 h the
reaction mixture was partitioned between ether (700 mL) and water (300 mL) and the separated aqueous
phase extracted with ether (3 × 400 mL). The combined organic phases where then washed with HCl (2 ×
100 mL of a 1 M aqueous solution) and brine (2 × 200 mL) before being dried (MgSO4), filtered and
concentrated under reduced pressure to afford acetate (21)16 (20.0 g, 100%) as a clear colorless oil, Rf 0.5
1
(7:93 v/v ethyl acetate/hexane) (Found: M+•, 168.1153. Calcd for C10H16O2 M+•, 168.1150). H NMR (300
MHz) δ 5.41 (m, 1H), 5.30 (m, 1H), 2.01 (s, 3H), 2.02–1.88 (complex m, 3H), 1.80–1.56 (complex m,
13
5H), 0.98 (t, J 7.5 Hz, 3H); C NMR (75 MHz) δ 170.8 (C), 146.2 (C), 118.1 (CH), 68.8 (CH), 30.2
(CH2), 28.3 (CH3), 28.2 (CH2), 21.4 (CH2), 19.1 (CH2), 11.8 (CH3); IR (neat) vmax 2937, 2876, 1732, 1667,
1456, 1370, 1312, 1242, 1163, 1022, 956, 910 cm–1; MS (EI) m/z 168 [(M+•, 2%], 126 (3), 109 (50), 108
(55), 97 (100), 93 (85), 79 (98), 67 (57).

84
HETEROCYCLES, Vol. 68, No. 1, 2006
2-(3-Ethylcyclohex-2-enyl)malonic Acid Dimethyl Ester (22).
1,2-Bis(diphenylphosphino)ethane (131 mg, 0.34 mmol) and Pd2dba3 (181 mg, 0.17 mmol) were added to
a magnetically stirred solution of acetate (21) (5.70 g, 29 mmol) in THF (50 mL). After 10 min a solution
of sodium dimethyl malonate in THF [prepared by the slow addition of malonic acid dimethyl ester (3.17
g, 49 mmol) to a magnetically stirred suspension of sodium hydride (1.18 g, 49 mmol) in THF (100 mL)]
was added dropwise. The resulting solution was heated at reflux for 24 h then cooled to 18°C, diluted
with ether (500 mL) and washed with water (1 × 200 mL). The separated aqueous phase was extracted
with ether (1 × 200 mL) and the combined ethereal fractions were then washed with brine (1 × 150 mL)
before being dried (MgSO4), filtered and concentrated under reduced pressure. Subjection of the ensuing
residue to flash chromatography (7:93 v/v ethyl acetate/hexane elution) afforded two fractions, A and B.
Concentration of fraction A (Rf 0.5) afforded the starting acetate (21) (1.50 g, 26% recovery) as a clear,
colorless oil and identical, in all respects, with an authentic sample.16
Concentration of fraction B (Rf 0.2) provided malonate (22) (5.42 g, 90% at 74% conversion) as a clear
1
colorless oil [Found: (M+H)+, 241.1437. C13H20O4 requires (M+H)+, 241.1440]. H NMR (300 MHz) δ
5.16 (s, 1H), 3.70 (s, 3H), 3.68 (s, 3H), 3.21 (d, J 9.6 Hz, 1H), 2.83 (br s, 1H), 1.96–1.82 (complex m,
3H), 1.74–1.43 (complex m, 3H), 1.30–1.18 (complex m, 2H), 0.92 (t, J 7.5 Hz, 3H); 13C NMR (75 MHz)
δ 168.7 (C), 168.6 (C), 142.1 (C), 119.6 (CH), 57.1 (CH), 52.2 (CH3), 35.5 (CH), 30.6 (CH2), 28.1 (CH2),
26.6 (CH2), 21.2 (CH2), 12.3 (CH3) (one signal obscured or overlapping); IR (neat) vmax 2933, 1738, 1435,
1333, 1296, 1245, 1193, 1152, 1022 cm–1; MS (EI) m/z 240 [(M)+•, 12%], 180 (100), 165 (15), 151 (50),
133 (42), 109 (75), 108 (69), 81 (65), 67 (48).
(3-Ethylcyclohex-2-enyl)acetic Acid Methyl Ester (23).
A magnetically stirred solution of malonate (22) (9.60 g, 41.2 mmol) in degassed DMSO (125 mL) was
treated, in one portion, with LiI•3H2O (7.75 g, 41.2 mmol). The ensuing mixture was heated at 180°C for
1.5 h then cooled to 18°C, diluted with water (75 mL) and extracted with ether (3 × 125 mL). The
combined organic fractions were then washed with brine (1 × 120 mL before being dried (MgSO4)
filtered and concentrated under reduced pressure. Subjection of the resulting dark-brown oil to flash
chromatography (5:95 → 7:93 v/v ethyl acetate/hexane gradient elution) afforded, after concentration of
the appropriate fractions (Rf 0.5 in 7:93 v/v ethyl acetate/hexane), the title methyl ester (23) (5.96 g, 80%)
as a clear, colorless oil (Found: M+•, 182.1304. C11H18O2 requires M+•, 182.1307). 1H NMR (300 MHz) δ
5.18 (s, 1H), 3.62 (s, 3H), 2.50 (br s, 1H), 2.20 (m, 2H), 2.00–1.08 (complex m, 8H), 0.93 (t, J 7.5 Hz,
13
3H); C NMR (75 MHz) δ 173.0 (C), 140.4 (C), 122.3 (CH), 51.2 (CH3), 40.9 (CH2), 32.3 (CH), 30.4
(CH2), 28.8 (CH2), 28.2 (CH2), 21.4 (CH2), 12.3 (CH3); IR (neat) vmax 2930, 2876, 1740, 1435, 1361, 1279,
1245, 1163, 1015, 884, 838 cm–1; MS (EI) m/z 182 (M+•, 40%), 167 (55), 153 (22), 122 (42), 109 (100),
108 (90), 93 (60), 79 (72), 67 (52).

HETEROCYCLES, Vol. 68, No. 1, 2006
85
(3-Ethylcyclohex-2-enyl)acetaldehyde (24).
A magnetically stirred solution of ester (23) (1.7 g, 9.39 mmol) in CH2Cl2 (42 mL) was cooled to –78°C
then DIBAL-H (10.33 mL of a 1 M solution in hexanes, 10.33 mmol) was added dropwise over 25 min.
After a further 1 h the reaction mixture was quenched by the addition of methanol (3 mL) then potassium
sodium tartrate (50 mL of a saturated aqueous solution). The ensuing mixture was extracted with ether (3
× 100 mL) and the combined organic fractions then washed with brine (1 × 100 mL) before being dried
(MgSO4), filtered and concentrated under reduced pressure to provide an opaque and light-yellow oil.
Subjection of this material to flash chromatography (7:93 v/v ethyl acetate/hexane elution) afforded, after
concentration of the appropriate fractions (Rf 0.5), the aldehyde (24) (1.09 g, 75%) as an opaque and
light-yellow oil (Found: M+•, 152.1202. C10H16O requires M+•, 152.1201). 1H NMR (300 MHz) δ 9.76 (s,
1H), 5.21 (s, 1H), 2.66 (br s, 1H), 2.38 (m, 2H), 1.98–1.45 (complex m, 6H), 1.28–1.12 (complex m, 2H),
0.96 (t, J 7.2 Hz, 3H); 13C NMR (75 MHz) δ 202.6 (CH), 140.9 (C), 122.0 (CH), 50.4 (CH2), 30.5 (CH2),
30.3 (CH), 29.1 (CH2), 28.1 (CH2), 21.5 (CH2), 12.3 (CH3); IR (neat) vmax 2962, 2928, 2714, 1725, 1456,
1263, 1117, 1056, 1025, 920, 840 cm–1; MS (EI) m/z 152 (M+•, 25%), 123 (75), 109 (45), 108 (55), 107
(45), 95 (37), 88 (60), 79 (58), 70 (82), 61 (100).
[2-(3-Ethylcyclohex-2-enyl)ethyl]prop-2-ynylamine (25).
A solution of aldehyde (24) (1.09 g, 7.21 mmol) in THF (30 mL) containing propargylamine (1.98 g, 36.1
mmol) and 4 Å molecular sieves (1.5 g) was stirred at 18°C for 16 h then filtered through a pad of dry
Celite™ and the filtrate concentrated under reduced pressure to a light-yellow oil presumed to contain the
expected imine. A solution of this material in methanol (20 mL) was cooled to 0°C then treated with
NaBH4 (301 mg, 7.93 mmol). After 0.33 h the reaction mixture was diluted with water (20 mL) and
extracted with ethyl acetate (3 × 60 mL). The combined organic fractions were washed with brine (1 × 50
mL) before being dried (MgSO4) filtered and concentrated under reduced pressure to give the title amine
(25) (1.03 g, 75%) as an opaque and light-yellow oil (Found: M+•, 191.1676. C13H21N requires M+•,
191.1674). 1H NMR (300 MHz) δ 5.24 (s, 1H), 3.42 (d, J 2.4 Hz, 2H), 2.73 (t, J 7.5 Hz, 2H), 2.20 (t, J 2.4
Hz, 1H), 2.10 (br s, 1H), 1.98–1.82 (complex m, 4H), 1.72 (m, 2H), 1.60–1.38 (complex m, 3H), 1.15 (m,
1H), 0.96 (t, J 7.5 Hz, 3H); 13C NMR (75 MHz) δ 138.8 (C), 123.4 (CH), 81.9 (C), 70.8 (CH), 45.9 (CH2),
37.7 (CH2), 36.2 (CH2), 32.8 (CH), 30.2 (CH2), 28.8 (CH2), 28.1 (CH2), 21.5 (CH2), 12.0 (CH3); IR (neat)
vmax 3308, 2925, 2854, 1678, 1612, 1455, 1371, 1327, 1118, 906, 840, 646 cm–1; MS (EI) m/z 191 (M+•,
10%), 190 (6), 176 (10), 163 (40), 162 (23), 136 (100), 121 (32), 107 (75), 93 (20), 79 (33), 68 (81). This
material was sufficiently pure for use in subsequent transformations.
N-Chloro-[2-(3-ethylcyclohex-2-enyl)ethyl]prop-2-ynylamine (11).
A magnetically stirred solution of amine (25) (100 mg, 0.52 mmol) in CH2Cl2 (4 mL) maintained at
–30°C was treated, in one portion, with N-chlorosuccinimide (77 mg, 0.58 mmol). The ensuing mixture

86
HETEROCYCLES, Vol. 68, No. 1, 2006
was warmed to 18°C over 1 h then the entire reaction mixture was added to the top of a pad of flash
chromatographic-grade silica containing in a sintered-glass funnel and the pad eluted with 3:7 v/v ethyl
acetate/hexane. Concentration of the relevant fractions of the eluent afforded the rather unstable
N-chloroamine (11) (118 mg, 100 %) as a pale-yellow oil, Rf 0.35 (1:9 v/v ethyl acetate/hexane). 1H NMR
(300 MHz) δ 5.23 (s, 1H), 3.82 (d, J 1.8 Hz, 2H), 3.05 (t, J 7.2 Hz, 2H), 2.41 (t, J 1.8 Hz, 1H), 2.15 (br s,
1H), 2.00–1.40 (complex m, 9H), 1.15 (m, 1H), 0.97 (t, J 7.5 Hz, 3H); 13C NMR (75 MHz) δ 140.2, 123.7,
77.3, 75.0, 60.3, 52.8, 35.0, 33.3, 30.9, 29.4, 28.8, 22.2, 12.8.
PTOC-Derivative (12) of [2-(3-Ethylcyclohex-2-enyl)ethyl]prop-2-ynylamine.
A solution of amine (25) (100 mg, 0.52 mmol) and triethylamine (58 mg, 0.57 mmol) in benzene (2 mL)
was added dropwise to a magnetically stirred suspension of 1-oxa-2-oxo-3-thiaindolizinium chloride18
(120 mg, 0.63 mmol) in benzene (2 mL). The ensuing mixture was stirred at 18°C for 2 h then filtered
through a pad of flash chromatographic-grade silica gel contained in a sintered-glass funnel. The pad was
then eluted with benzene and the yellow-colored fractions concentrated under reduced pressure to yield
the title compound (12) (100 mg, 60%) as a clear, yellow oil. This highly unstable material was used,
without further purification, in the relevant radical cyclization reaction detailed in Section B.
N-Allyl-[2-(3-ethylcyclohex-2-enyl)ethyl]amine (26).
A magnetically stirred solution of aldehyde (24) (190 mg, 1.25 mmol) in THF (5 mL) was treated with
allylamine (350 mg, 6.25 mmol) followed by activated 4 Å molecular sieves (250 mg). The resulting
mixture was stirred at 18°C for 4 h then filtered through a pad of dry Celite™ contained in a
sintered-glass funnel. The retained solids were washed with THF (10 mL) and the combined filtrates then
concentrated under reduced pressure to give a light-yellow oil. A magnetically stirred solution of this
material in methanol (3 mL) was cooled to 0°C and treated, in one portion, with NaBH4 (47.5 mg, 1.25
mmol). After 0.3 h the reaction mixture was diluted with water (10 mL) and extracted with ethyl acetate
(3 × 40 mL). The combined organic fractions were washed with brine (1 × 30 mL) then dried (MgSO4),
filtered and concentrated under reduced pressure to give the title compound (26) (181 mg, 75%) as an
opaque oil (Found: M+•, 193.1830. C13H23N requires M+•, 193.1831). 1H NMR (300 MHz) δ 5.86 (m, 1H),
5.21 (s, 1H), 5.12 (dm, J 17.1 Hz, 1H), 5.03 (dm, J 10.2 Hz, 1H), 3.21 (d, J 6.0 Hz, 2H), 2.62 (t, J 7.8 Hz,
2H), 2.08 (br s, 1H), 1.94−1.82 (complex m, 4H), 1.68 (m, 2H), 1.51–1.34 (complex m, 4H), 1.10 (m,
1H), 0.93 (t, J 7.2 Hz, 3H); 13C NMR (75 MHz) δ 139.3 (C), 136.7 (CH), 123.8 (CH), 115.5 (CH2), 52.5
(CH2), 47.1 (CH2), 36.9 (CH2), 33.3 (CH), 30.5 (CH2), 29.1 (CH2), 28.4 (CH2), 21.8 (CH2), 12.4 (CH3); IR
(neat) vmax 2962, 2926, 2856, 1643, 1455, 1371, 1262, 1117, 993, 916, 839, 743 cm–1; MS (EI) m/z 193
(M+•, 21%), 178 (5), 164 (16), 150 (10), 136 (77), 121 (25), 107 (65), 79 (31), 70 (100).
N-Allyl-N-chloro-[2-(3-ethylcyclohex-2-enyl)ethyl]amine (13).
A magnetically stirred solution of amine (26) (100 mg, 0.52 mmol) in CH2Cl2 (4 mL) was cooled to

HETEROCYCLES, Vol. 68, No. 1, 2006
87
–30°C then treated with N-chlorosuccinimide (83 mg, 0.62 mmol) and the resulting mixture allowed to
warm to 18°C over 2 h then loaded onto the top of a pad of flash chromatographic-grade silica gel
contained in a sintered-glass funnel. The pad was eluted with 1:1 v/v ethyl acetate/hexane and
concentration of the appropriate fractions (Rf 0.8 in 1:4 v/v ethyl acetate/hexane) under reduced pressure
1
then afforded the title compound (13) (114 mg, 97%) as an unstable and pale-yellow oil. H NMR (300
MHz) δ 5.96 (m, 1H), 5.32–5.22 (complex m, 3H), 3.61 (dt, J 6.6 and 1.5 Hz, 2H), 2.99 (t, J 7.5 Hz, 2H),
2.15 (br s, 1H), 2.00–1.86 (complex m, 3H), 1.79–1.44 (complex m, 6H), 1.15 (m, 1H), 0.99 (t, J 7.5 Hz,
3H); 13C NMR (75 MHz) δ 139.7 (C), 133.5 (CH), 123.5 (CH), 119.0 (CH2), 66.7 (CH2), 60.9 (CH2), 34.5
(CH2), 33.0 (CH), 30.6 (CH2), 29.1 (CH2), 28.4 (CH2), 21.8 (CH2), 12.4 (CH3); IR (neat) vmax 2927, 1374,
1295, 1184, 922, 817, 647 cm–1. Satisfactory MS spectral data could not be obtained on this material.
But-3-ynyl-[2-(3-ethylcyclohex-2-enyl)ethyl]amine (27).
A solution of triphenylphosphine (176 mg, 0.6 mmol) in THF (4 mL) was added to a magnetically stirred
solution of homopropargyl azide22 (57 mg, 0.6 mmol) in THF (4 mL). After 2 h aldehyde (24) (91 mg, 0.6
mmol) was added to the reaction mixture and the ensuing solution heated at reflux for 16 h. The reaction
mixture was then cooled to 0°C and NaCNBH3 (57 mg, 0.9 mmol) added. After 0.5 h the reaction mixture
was poured into NaOH (50 mL of a 2 M aqueous solution) and extracted with ether (3 × 75 mL). The
organic phases were combined, washed with brine (1 × 100 mL), then dried (MgSO4), filtered and
concentrated under reduced pressure to give a light-yellow oil. This material was subject to flash
chromatography (74:25:1 v/v/v ethyl acetate/hexane/triethylamine elution) and thereby yielding, after
evaporation of appropriate fractions (Rf 0.1 in 3:1 v/v ethyl acetate/hexane), homopropargylamine (27)
1
(40 mg, 33%) as a clear, colorless oil (Found: M+•, 205.1830. C14H23N requires M+•, 205.1831). H NMR
(300 MHz) δ 5.25 (s, 1H), 2.79 (t, J 6.6 Hz, 2H), 2.68 (t, J 7.5 Hz, 2H), 2.40 (m, 2H), 2.11 (m, 1H),
2.01–1.84 (complex m, 4H), 1.73 (m, 2H), 1.58–1.40 (complex m, 4H), 1.25–1.09 (complex m, 2H), 0.97
13
(t, J 7.5 Hz, 3H); C NMR (75 MHz) δ 139.5 (C), 123.8 (CH), 82.4 (C), 69.5 (CH), 48.0 (CH2), 47.1
(CH2), 36.8 (CH2), 33.4 (CH), 30.6 (CH2), 29.2 (CH2), 28.5 (CH2), 21.9 (CH2), 19.5 (CH2), 12.5 (CH3); IR
(neat) vmax 2956, 2920, 2847, 1454, 1371, 1122, 1078, 1053, 1035 630 cm–1; MS (EI) m/z 205 (M+•, 10%),
166 (22), 136 (100), 121 (12), 107 (63), 82 (25), 79 (28), 67 (18).
PTOC-Derivative (14) of But-3-ynyl-[2-(3-ethylcyclohex-2-enyl)ethyl]amine.
A solution of amine (27) (10 mg, 0.05 mmol) and triethylamine (5.0 mg, 0.05 mmol) in benzene (0.5 mL)
was added dropwise to a magnetically stirred suspension of 1-oxa-2-oxo-3-thiaindolizinium chloride18 (11
mg, 0.06 mmol) in benzene (0.5 mL). The ensuing mixture was stirred at 18°C for 2 h then filtered
through a pad of flash chromatographic-grade silica gel contained in a sintered-glass funnel. The pad was
then eluted with benzene. Concentration of the ensuing yellow-colored fractions under reduced pressure
afforded the title compound (14) (17 mg, quant.) as a clear, yellow oil. This instability of this material

88
HETEROCYCLES, Vol. 68, No. 1, 2006
precluded the acquisition of spectroscopic data and it was used, without purification, in the radical
cyclization reaction detailed in Section B.
N-Chloro-N-[2-(3-ethylcyclohex-2-enyl)ethyl]but-3-ynylamine (15).
A magnetically stirred solution of amine (27) (40 mg, 0.20 mmol) in CH2Cl2 (4 mL) was cooled to –30°C
then treated with N-chlorosuccinimide (29 mg, 0.22 mmol). The resulting mixture was allowed to warm
to 18°C over 1 h then adsorbed onto a pad of flash chromatographic-grade silica contained in a
sintered-glass funnel. The pad was then eluted with 3:7 v/v ethyl acetate/hexane. Concentration of the
relevant fractions (Rf 0.5) then afforded the unstable N-chloroamine (15) (42 mg, 92%) as a pale-yellow
oil. 1H NMR (300 MHz) δ 5.24 (s, 1H), 3.11 (t, J 7.5 Hz, 2H), 3.03 (t, J 7.5 Hz, 2H), 2.59 (m, 2H), 2.11
(br s, 1H), 2.01–1.85 (complex m, 4H), 1.80–1.43 (complex m, 6H), 1.15 (complex m, 1H), 0.97 (t, J 7.5
Hz, 3H).
Section B: Radical Cyclization Studies
1-Methylenedecahydropyrolo[3,2,1-hi]indole (28).
A magnetically stirred solution of N-chloroamine (10) (31 mg, 0.15 mmol) and n-Bu3SnH (54 mg, 0.15
mmol) in benzene (15.7 mL) was treated with AIBN (3 mg, 5 mol%) and the ensuing mixture heated at
reflux for 1.5 h then cooled to 18°C and extracted with HCl (1 × 50 mL of a 3 M aqueous solution). The
combined aqueous phases were basified with NaOH (ca. 100 mL of a 2 M solution) then extracted with
ethyl acetate (3 × 50 mL). The combined organic phases were then dried (MgSO4), filtered and
concentrated under reduced pressure to give a light-yellow oil. Subjection of this material to flash
chromatography (50:2:2:1 v/v/v/v ethyl acetate/methanol/triethylamine/water) and concentration of the
appropriate fractions (Rf 0.2) afforded the title amine (28) (19 mg, 74%) as a pale-yellow oil (Found: M+•,
1
163.1354. C11H17N requires M+•, 163.1361). H NMR (300 MHz) δ 4.81 (s, 1H), 4.78 (s, 1H), 3.56 (m,
13
1H), 3.21 (dq, J 13.3 and 1.5 Hz, 1H), 3.02 (m, 1H), 2.57 (m, 1H), 2.15–1.50 (complex m, 10H); C
NMR (75 MHz) δ 155.6 (C), 103.5 (CH2), 65.5 (CH), 58.7 (CH2), 54.3 (CH2), 41.2 (CH), 35.7 (CH), 34.4
(CH2), 28.5 (CH2), 28.4 (CH2), 20.3 (CH2); IR (neat) vmax 2929, 2855, 1665, 1614, 1447, 1288, 1144, 1111,
883 cm–1; MS (EI) m/z 163 (M+•, 45%), 162 (100), 120 (70), 82 (30), 68 (100).
8a-Ethyl-1-methylenedecahydropyrolo[3,2,1-hi]indole (29).
A magnetically stirred solution of amine (11) (70 mg, 0.31 mmol) and n-Bu3SnH (90 mg, 0.31 mmol) in
toluene (5 mL) was treated with AIBN (10 mg, 20 mol %) then heated at reflux for 5 h. The cooled
mixture was extracted with HCl (1 × 100 mL of a 2 M aqueous solution) and the separated aqueous phase
basified using NaOH (ca. 110 mL of a 2 M aqueous solution) then extracted with ether (3 × 100 mL). The
combined organic extracts were washed with brine (1 × 100 mL) before being dried (MgSO4), filtered and
concentrated under reduced pressure to afford a yellow oil. Subjection of this material to flash
chromatography (basic alumina, benzene → 1:3:6 v/v/v methanol/ethyl acetate/benzene gradient elution)

HETEROCYCLES, Vol. 68, No. 1, 2006
89
afforded, after concentration of the relevant fractions (Rf = 0.1 in 1:4 v/v methanol/ethyl acetate, silica),
compound (29) (40 mg, 68%) as a pale-yellow and opaque oil [Found: (M+H)+, 192.1748. C13H21N
requires (M+H)+, 192.1752]. 1H NMR (300 MHz) δ 4.85 (t, J 1.8 Hz, 1H), 4.63 (t, J 1.8 Hz, 1H), 3.82 (dt,
J 14.4 and 1.8 Hz, 1H), 3.27 (d, J 7.2 Hz, 1H), 3.24 (dt, J 14.4 and 1.8 Hz, 1H), 3.01 (dt, J 10.5 and 6.3
Hz, 1H), 2.66 (dt, J 10.5, and 6.0 Hz, 1H), 2.07 (m, 1H), 1.82 (m, 1H), 1.66–1.30 (complex m, 9H), 0.85
(t, J 7.2 Hz, 3H); 13C NMR (75 MHz) δ 155.7 (C), 102.5 (CH2), 72.0 (CH), 61.5 (CH2), 55.2 (CH2), 46.2
(C), 36.8 (CH), 33.9 (CH2), 33.0 (CH2), 32.0 (CH2), 26.6 (CH2), 18.5 (CH2), 8.8 (CH3); MS (EI) m/z 192
[(M+H)+, 5%], 191 (M+•, 22), 190 (100), 162 (28), 120 (45), 68 (60).
8a-Ethyl-1-(pyridin-2-ylsulfanylmethylene)decahydropyrrolo[3,2,1-hi]indole (30).
A magnetically stirred solution of compound (12) (100 mg, 0.30 mmol) in acetonitrile (6 mL) containing
malonic acid (93 mg, 0.90 mmol) or BF3•Et2O (126 mg, 0.90 mmol) was irradiated with a 250 W tungsten
lamp for 0.5 h then cooled and partitioned between ether (100 mL) and HCl (150 mL of a 3 M aqueous
solution). The aqueous phase was then basified with NaOH (ca. 200 mL of a 2 M aqueous solution),
extracted with ethyl acetate (3 × 100 mL) and the combined organic phases then washed with brine (1 ×
10 mL) before being dried (MgSO4), filtered and concentrated under reduced pressure to give the title
compound (30) [118 mg, 65% (when malonic acid was used) or 128 mg, 71% (when BF3•Et2O was used)]
1
as a pale-orange oil (Found: M+•, 300.1660. C18H24N2S requires M+•, 300.1660). H NMR (300 MHz) δ
8.42 (dm, J 4.8 Hz, 1H), 7.50 (td, J 7.5 and 2.1 Hz, 1H), 7.13 (dm, J 7.5 Hz, 1H), 6.99 (m, 1H), 6.25 (t, J
2.1 Hz, 1H), 3.94 (dd, J 16.2 and 2.1 Hz, 1H), 3.40 (dd, J 16.2 and 2.1 Hz, 1H), 3.31 (d, J 7.5 Hz, 1H),
3.02 (dt, J 10.8 and 6.6 Hz, 1H), 2.72 (dt, J 10.8 and 6.3 Hz, 1H), 2.08 (m, 1H), 1.78 (m, 1H), 1.66–1.38
13
(complex m, 9H), 0.90 (t, J 7.5 Hz, 3H); C NMR (75 MHz) δ 158.6 (C), 151.8 (C), 149.5 (CH), 136.2
(CH), 121.2 (CH), 119.7 (CH), 106.4 (CH), 72.7 (CH2), 60.6 (CH2), 55.6 (CH), 48.4 (C), 36.8 (CH), 35.3
(CH2), 32.4 (CH2), 31.0 (CH2), 26.0 (CH2), 18.3 (CH2), 8.9 (CH3); MS (EI) m/z 300 (M+•, 2%), 299 (1),
267 (1), 220 (7), 191 (32), 190 (100), 160 (12).
8a-Ethyl-1-methyldecahydropyrrolo[3,2,1-hi]indole (31).
A magnetically stirred solution of amine (13) (210 mg, 0.93 mmol) and n-Bu3SnH (323 mg, 1.12 mmol)
in benzene (100 mL) was treated with AIBN (30 mg, 20 mol%) and the ensuing mixture heated at reflux
for 6 h then cooled and extracted with HCl (200 mL of a 3 M solution). The combined aqueous phases
were basified with NaOH (ca. 300 mL of a 2 M aqueous solution) then extracted with ethyl acetate (3 ×
100 mL). The combined organic phases were dried (MgSO4), filtered and concentrated under reduced
pressure to give a light-yellow oil. Subjection of this material to flash chromatography (basic alumina,
benzene → 3:7 v/v ethyl acetate/benzene → 5:30:65 v/v/v methanol/ethyl acetate/benzene gradient
elution) afforded, after concentration of the appropriate fractions (Rf = 0.1 in 1:4 v/v methanol/ethyl
acetate, silica), the title amine (31) (130 mg, 62%) as a clear, colorless oil (Found: M+•, 193.1835.

90
HETEROCYCLES, Vol. 68, No. 1, 2006
1
C13H23N requires M+•, 193.1831). H NMR (300 MHz) δ 3.29 (d, J 6.6 Hz, 1H), 3.12 (m, 1H), 2.90 (dd, J
9.6 and 2.4 Hz, 1H), 2.81 (dd, J 9.6 and 5.4 Hz, 1H), 2.56 (m, 1H), 2.16–1.95 (complex m, 3H),
1.76–1.62 (complex m, 3H), 1.52–1.11 (complex m, 6H), 1.03 (d, J 7.2 Hz, 3H), 0.84 (t, J 7.5 Hz, 3H);
13C NMR (75 MHz) δ 70.7, 61.1, 53.0, 44.8, 43.7, 35.5, 34.6, 30.5, 28.9, 25.2, 19.2, 14.9, 9.0; IR (neat)
vmax 2928, 1452, 1379, 1261, 1151, 799 cm–1; MS (EI) m/z 193 (M+•, 32%), 192 (100), 164 (12), 136 (10),
124 (12), 122 (15), 96 (12), 82 (31), 70 (25), 55 (15), 41 (24).
Attempted Cyclization of Compound 14.
A magnetically stirred solution of compound (14) (17 mg, 0.05 mmol) in acetonitrile (2 mL) containing
malonic acid (15 mg, 0.14 mmol) or BF3•Et2O (21 mg, 0.14 mmol) was irradiated with a 250 W tungsten
lamp for 0.5 h then cooled and partitioned between ether (20 mL) and HCl (20 mL of a 3 M aqueous
solution). The separated aqueous phase was extracted with ethyl acetate (2 × 20 mL) and the combined
organic phases then washed with brine (1 × 5 mL) before being dried (MgSO4), filtered and concentrated
under reduced pressure to give a light-yellow oil (10 mg). 1H NMR spectroscopic analysis of this material
suggested that extensive decomposition of the starting material had occurred.
Attempted Cyclization of Compound 15.
A magnetically stirred solution of N-chloroamine (15) (27 mg, 0.11 mmol) and n-Bu3SnH (33 mg, 0.11
mmol) in toluene (4 mL) was treated with AIBN (3 mg, 20 mol%) and the ensuing mixture heated at
reflux for 1 h then cooled to 18°C and extracted with HCl (20 mL of a 3 M aqueous solution). The
combined aqueous phases were basified with NaOH (ca. 30 mL of a 2 M aqueous solution) and extracted
with ethyl acetate (3 × 10 mL). The combined organic phases were then dried (MgSO4), filtered and
1
concentrated under reduced pressure to give a yellow oil. Analysis of this oil by H NMR spectroscopy
indicated that only reductive dechlorination of compound (15) had occurred and thus producing amine
(27) (ca. 10 mg, ca. 50%).
ACKNOWLEDGMENTS
We thank the Australian Research Council and Institute of Advanced Studies for financial support
including the provision of a PhD scholarship to DWL.
REFERENCES
1. For reviews of various aspects of the chemistry and biochemistry of these compounds see (a) J.
Hájíček, Collect. Czech. Chem. Commun., 2004, 69, 1681 and references cited therein; (b) J. E.
Saxton, The Alkaloids, 1998, 51, 2; (c) J. E. Saxton, The Alkaloids, 1998, 50, 343; (d) T. M. Kutchan,
The Alkaloids, 1998, 50, 257; (e) R. Verpoorte, R. van der Heijden, and P. R. H. Moreno, The
Alkaloids, 1997, 49, 221.

HETEROCYCLES, Vol. 68, No. 1, 2006
91
2. (a) J. W. Moncrief and W. N. Lipscomb, J. Am. Chem. Soc., 1965, 87, 4963; (b) M. Gorman, N.
Neuss, and K. Biemann, J. Am. Chem. Soc., 1962, 84, 1058.
3. A. Brossi and M. Suffness, The Alkaloids, 1990, 37, 1.
4. For examples of recent efforts to prepare “medicinally active” vindoline derivatives see (a) M.
Fekete, P. Kolonits, and L. Novak, Heterocycles, 2005, 65, 165; (b) J. Fahy, V. Thillaye de Boullay,
and D. C. H. Bigg, Bioorg. Med. Chem. Lett., 2002, 12, 505.
5. (a) M. E. Kuehne and J. C. Bohnert, J. Org. Chem., 1981, 46, 3443; (b) C. Kan, H.-P. Husson, H.
Jacquemin, S.-K. Kan, and M. Lounasmaa, Tetrahedron Lett., 1980, 21, 55; (c) F. Khuong-Huu, M.
Cesario, J. Guilhem, and R. Goutarel, Tetrahedron, 1976, 32, 2539.
6. (a) M. E. Kuehne and J. C. Bohnert, J. Org. Chem., 1981, 46, 3443; (b) C. Kan, H.-P. Husson, S.-K.
Kan, and M. Lounasmaa, Tetrahedron Lett., 1980, 21, 3363.
7. (a) S.-z. Zhou, S. Bommezijn, and J. A. Murphy, Org. Lett., 2002, 4, 443; (b) S. Kobayashi, T. Ueda,
and T. Fukuyama, Synlett, 2000, 883; (c) M. E. Kuehne, D. E. Podhorez, T. Mulamba, and W. G.
Bornmann, J. Org. Chem., 1987, 52, 347; (d) P. L. Feldman and H. Rapoport, J. Am. Chem. Soc.,
1987, 109, 1603; (e) B. Danieli, G. Lesma, G. Palmisano, and R. Riva, J. Chem. Soc., Perkin Trans.
1, 1987, 155; (f) R. Z. Andriamialisoa, N. Langlois, and Y. Langlois, J. Org. Chem., 1985, 50, 961;
(g) B. Danieli, G. Lesma, G. Palmisano, and R. Riva, J. Chem. Soc., Chem. Commun., 1984, 909; (h)
J. P. Kutney, U. Bunzli-Trepp, K. K. Chan, J. P. de Souza, Y. Fujise, T. Honda, J. Katsube, F. K.
Klein, A. Leutwiller, S. Morehead, M. Rohr, and B. R. Worth, J. Am. Chem. Soc., 1978, 100, 4220;
(i) Y. Ban, Y. Sekine, and T. Oishi, Tetrahedron Lett., 1978, 19, 151; (j) S. Takano, K. Shishido, J.-i.
Matsuzaka, M. Sato, and K. Ogasawara, Heterocycles, 1979, 13, 307; (k) M. Ando, G. Büchi, and T.
Ohnuma, J. Am. Chem. Soc., 1975, 97, 6880.
8. M. E. Kuehne, U. K. Bandarage, A. Hammach, Y.-L. Li, and T. Wang, J. Org. Chem., 1998, 63,
2172.
9. (a) M. G. Banwell, D. A. S. Beck, and A. C. Willis, ARKIVOC, 2006, (iii), 163; (b) M. G. Banwell,
D. W. Lupton, and A. C. Willis, Aust. J. Chem., 2005, 58, 722; (c) M. G. Banwell and D. W. Lupton,
Org. Biomol. Chem., 2005, 3, 213; (d) M. G. Banwell, A. J. Edwards, K. A. Jolliffe, and J. A. Smith,
Org. Biomol. Chem., 2003, 1, 296; (e) M. G. Banwell and J. A. Smith, J. Chem. Soc., Perkin Trans.
1, 2002, 2613.
10. M. G. Banwell, A. J. Edwards, D. W. Lupton, and G. Whited, Aust. J. Chem., 2005, 58, 14.
11. J. Cassayre, F. Gagosz, and S. Z. Zard, Angew. Chem., Int. Ed., 2002, 41, 1783.
12. J. Cossy, L. Tresnard, and D. Gomez Pardo, Eur. J. Org. Chem., 1999, 1925.
13. For some further examples of key work in this area see: (a) H. Senboku, H. Hasegawa, K. Orito, and
M. Tokuda, Tetrahedron Lett., 2000, 41, 5699; (b) X. Lin, D. Stien, and S. M. Weinreb, Tetrahedron

92
HETEROCYCLES, Vol. 68, No. 1, 2006
Lett., 2000, 41, 2333; (c) A. J. Clark, R. J. Deeth, C. J. Samuel, and H. Wongtap, Synlett, 1999, 444;
(d) A. J. Clark, R. P. Filik, J. L. Paecock, and G. H. Thomas, Synlett, 1999, 441; (e) H. Senboku, H.
Hasegawa, K. Orito, and M. Tokuda, Heterocycles, 1999, 50, 333; (f) C. Ha, O. M. Musa, F. N.
Martinez, and M. Newcomb, J. Org. Chem., 1997, 62, 2704; (g) J. Boivin, A.-C. Callier-Dublanchet,
B. Quiclet-Sire, A.-M. Schiano, and S. Z. Zard, Tetrahedron, 1995, 51, 6517; (h) M. Tokuda, H.
Fujita, and H. Suginome, J. Chem. Soc., Perkin Trans. 1, 1994, 777; (i) W. R. Bowman, D. N. Clark,
and R. J. Marmon, Tetrahedron, 1994, 50, 1275; (j) A. L. J. Beckwith, B. J. Maxwell, and J.
Tsanaktsidis, Aust. J. Chem., 1991, 44, 1809; (k) M. Newcomb, D. J. Marquardt, and M. U. Kumar,
Tetrahedron, 1990, 46, 2345; (l) M. Newcomb, D. J. Marquardt, and T. M. Deeb, Tetrahedron, 1990,
46, 2329; (m) M. Newcomb, T. M. Deeb, and D. J. Marquardt, Tetrahedron, 1990, 46, 2317; (n) C.
A. Broka and J. F. Gerlits, J. Org. Chem., 1988, 53, 2144; (o) L. Stella, Angew. Chem., Int. Ed. Engl.,
1983, 22, 337; (p) J.-L. Bougeois, L. Stella, and J.-M. Surzur, Tetrahedron Lett., 1981, 22, 61; (q) R.
A. Perry, S. C. Chen, B. C. Menon, K. Hanaya, and Y. L. Chow, Can. J. Chem., 1976, 54, 2385; (r)
J.-M. Surzur, L. Stella, and P. Tordo, Tetrahedron Lett., 1970, 3107.
14. L. A. Paquette and G. Zon, J. Am. Chem. Soc., 1974, 96, 203.
15. B.-D. Chong, Y.-I. Ji, S.-S. Oh, J.-D. Yang, W. Baik, and S. Koo, J. Org. Chem., 1997, 62, 9323.
16. M. Hirano and T. Morimoto, J. Chem. Soc., Perkin Trans. 2, 1985, 1105.
17. E. Wenkert, T. D. J. Halls, G. Kunesch, K. Orito, R. L. Stephens, W. A. Temple, and J. S. Yadav, J.
Am. Chem. Soc., 1979, 101, 5370.
18. D. H. R. Barton, D. Crich, and W. B. Motherwell, Tetrahedron, 1985, 41, 3901.
19. For earlier applications of this protocol in related contexts see J. Mulzer, A. Meier, J. Buschmann,
and P. Luger, Synthesis, 1996, 123.
20. M. E. Kuehne, T. Wang, and P. J. Seaton, J. Org. Chem., 1996, 61, 6001.
21. A. L. J. Beckwith, Tetrahedron, 1981, 37, 3073.
22. E. C. Taylor, J. E. Macor, and J. L. Pont, Tetrahedron, 1987, 43, 5145.