Development of an intramolecular aryne ene reaction and application to the formal synthesis of (±)-crinine

Article abstract of DOI:10.1021/ja306881u

A general and high yielding annulation strategy for the synthesis of various carbo- and heterocycles, based on an intramolecular aryne ene reaction is described. It was found that the geometry of the olefin is crucial to the success of the reaction, with exclusive migration of the trans-allylic-H taking place. Furthermore, the electronic nature of the aryne was found to be important to the success of the reaction. Deuterium labeling studies and DFT calculations provided insight into the reaction mechanism. The data suggests a concerted asynchronous transition state, resembling a nucleophilic attack on the aryne. This strategy was successfully applied to the formal synthesis of the ethanophenanthridine alkaloid (±)-crinine.

Full text of DOI:10.1021/ja306881u

Article
pubs.acs.org/JACS
Development of an Intramolecular Aryne Ene Reaction and
Application to the Formal Synthesis of ( )-Crinine
David A. Candito, Dennis Dobrovolsky, and Mark Lautens*
Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, Ontario M5S 3H6, Canada
S
* Supporting Information
ABSTRACT: A general and high yielding annulation strategy
for the synthesis of various carbo- and heterocycles, based on
an intramolecular aryne ene reaction is described. It was found
that the geometry of the olefin is crucial to the success of the
reaction, with exclusive migration of the trans-allylic-H taking
place. Furthermore, the electronic nature of the aryne was
found to be important to the success of the reaction.
Deuterium labeling studies and DFT calculations provided
insight into the reaction mechanism. The data suggests a
concerted asynchronous transition state, resembling a nucleophilic attack on the aryne. This strategy was successfully applied to
the formal synthesis of the ethanophenanthridine alkaloid ( )-crinine.
INTRODUCTION
■
Arynes are among the first reactive intermediates studied by
organic chemists. Since their discovery they have fascinated
chemists from both theoretical and synthetic perspectives.
Although the field of aryne chemistry is relatively developed
and may be considered mature, it has recently undergone a
renaissance, as evidenced by the large number of reports in the
literature.1 This resurgence of aryne chemistry is partly due to
new and milder methods of aryne generation2 that permit new
reactions to be uncovered.
The aryne ene reaction, although known, has not gained
widespread use in synthetic organic chemistry, and only a few
limited reports are found in the literature. Often the authors
were studying other reactivity and observe an ene reaction as a
side pathway.3 Only recently have a few reports emerged that
utilize the aryne ene reaction as a synthetic methodology.4
Thus, the ene reaction warrants investigation as a methodology
in its own right.
Figure 1. Intramolecular aryne ene reaction.
heterocycles and demonstrate its application to the formal
synthesis of ( )-crinine.6
RESULTS AND DISCUSSION
■
Optimization Studies. Initial experiments began with
substrate 1.1 at very high dilution in order to minimize any
potential intermolecular reactions with the base or substrate.
Deprotonation was carried out at cryogenic temperatures (−78
°C) by slow addition of LDA using a syringe pump. Following
complete addition, the reaction was allowed to slowly warm to
room temperature. The reaction was found to be very sluggish
when carried out in this manner; however, 1.1 had cleanly
converted to 2.1. Increasing the concentration steadily
increased the conversion and yield (Table 1, entries 1−3),
and adding the base at room temperature at the beginning of
the reaction further improved the conversion (Table 1, entry
4). The improved conversion is likely a result of the more rapid
rate of LDA addition and the longer time spent at room
temperature. When 2 equiv of LDA was added at the beginning
of the reaction, 2.1 reacted further to form product 4.1, which
likely arose from deprotonation at the benzylic position of 2.1
We postulated that the lack of literature reports on the
subject was related to the difficulty in controlling the chemo-
and regioselectivity of the ene reaction, rather than a lack of
interest. The intramolecular Diels−Alder reaction of arynes has
been used as a successsful strategy in organic synthesis, and
drawing inspiration from the work of Martin,5a Buszek,5d
Danhieser,5e and Castedo and Guitan 5b,c
who have demon-
́
,
strated very elegant examples of these reactions, we wondered
whether the outcome of the reaction could be controlled by
tethering the olefin to the aryne precursor (Figure 1). Doing so
would allow for the use of deprotonation-based methods of
aryne generation, thereby simplifying the starting materials
employed and providing an interesting annulation strategy.
Herein, we report the development of a general, high
yielding, and selective intramolecular aryne ene reaction
providing straightforward access to benzofused carbo- and
Received: July 20, 2012
Published: August 20, 2012
© 2012 American Chemical Society
15572
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580

Journal of the American Chemical Society
Article
concerted manner;9 however, a radical mechanism has also
been proposed.9d Another possibility would be a polar
mechanism where the olefin attacks the aryne and a zwitterionic
intermediate is formed. Clearly each of these mechanisms will
possess different structural requirements of the starting
material. Even a partial understanding of the mechanism
would greatly aid the use of this reaction in organic synthesis.
We decided to probe the mechanism using deuterium
labeling experiments. Reaction of labeled substrate 1.1-d6 led to
complete transfer of deuterium (Scheme 2), consistent with a
a
Table 1. Optimization
b
b
conc.
(M)
conv.
(%)
2.1
3.1
4.1
entry
base
(%)
d
(%)
(%)
c
c
c
,
e
1
2
3
4
5
6
7
8
9
LDA
LDA
LDA
LDA
LDA
0.003
0.012
0.024
0.024
0.024
0.024
0.024
0.024
0.024
23
54
23
52
63
64
Scheme 2. Reactions of Deuterium Labeled Substrates
d
100
100
<5
80
88
<5
d
f
54
40
LiHMDS
KHMDS
67
39
23
d
LiTMP
77
38
d
Me2Zn(TMP)Li
51
28
a
Reaction conditions: 1.1 (1 equiv, 1 mmol), base (1.1 equiv), THF
b
1
(conc. based on substrate), rt, 24 h. Determined by H NMR using
mesitylene as an internal standard. Lithium amide added at −78 °C
using syringe pump (rate = 6 mL/h, LDA 0.1 M), then rt, 24 h (total
c
unimolecular concerted process. Inspection of hand-held and
computer models suggested that the olefin geometry was
important for the reaction and there should be a preference for
migration of the trans-allylic-H. In order to test this hypothesis,
substrate 1.1-d3 was synthesized with deuteration at the trans-
methyl group, as a 91:9 mixture of isomers. Reaction of
substrate 1.1-d3 led to complete deuterium transfer, which
confirmed the preference for the trans-allylic-H to migrate and
further reinforced the notion that no radical or polar
intermediates were formed in this reaction, because rotation
about the single bond of the intermediate radical or cation
would have led to scrambling of the label.
We were intrigued by the complete and selective deuterium
transfer from 1.1-d3 and wondered whether this was a kinetic
effect (i.e., the trans-substituent reacts much faster than the cis-
substituent) or whether cis-olefins are incapable of reacting. In
order to test this hypothesis, we synthesized substrates cis- and
trans-1.2 and subjected each to the reaction conditions. The
substrates displayed marked differences in reactivity (Table 2).
d
e
f
time). Isolated yield. 48 h. LDA (2 equiv).
and subsequent ring-opening (Table 1, entry 5). Other lithium
amide bases were also investigated; LiHMDS did not lead to
any conversion, likely due to its decreased basicity, and
KHMDS was a competent base but did not prove advantageous
(Table 1, entries 6 and 7). Interestingly, LiTMP and
Me2Zn(TMP)Li2b gave mixtures of 2.1 and 3.1, which likely
arose by deprotonation of 1.1 at the benzylic position and
subsequent [2,3]-Wittig rearrangement.7 The difference in
reactivity between LDA, LiTMP, and Me2Zn(TMP)Li is likely
due to the difference in the steric properties of these bases as
opposed to the difference in pKa; the larger LiTMP and
Me2Zn(TMP)Li complexes deprotonate the more accessible
benzylic protons as opposed to the aromatic proton. With
suitable conditions identified for the intramolecular aryne ene
reaction, we went on to investigate both the scope and
mechanism.
Mechanistic Studies. Early on, we were interested in
probing the mechanism of this reaction and determining the
critical factors necessary for its success. When the leaving group
is a bromide, the deprotonation by LDA is likely to be the rate-
determining step of the reaction1a,8 because elimination is fast
and because of the low concentrations employed. The
deprotonation is expected to initially generate a lithiated
intermediate, which undergoes elimination to generate the
aryne intermediate. The aryne then engages the tethered olefin
in an ene reaction to give the final product (Scheme 1). The
aryne ene reaction is generally believed to proceed in a
a
Table 2. Reactions of cis- and trans-1.2
b
b
c
entry
1.2
cis/trans conv. (%) ratio (2.30/5.1) yield (%)
1
2
trans-1.2
cis-1.2
6:94
98:2
79
84
94:6
2:98
69
73
a
Reaction conditions: 1.2 (1 equiv, 0.5 mmol), LDA (1.1 equiv), THF
b
1
(0.024 M, based on substrate), rt, 2 h. Determined by H NMR
analysis of the crude reaction mixture. Combined isolated yield.
c
Scheme 1. Proposed Mechanism
The compound trans-1.2 reacted to give 2.30 as the major
product (entry 1), whereas cis-1.2 reacted to give 5.1 as the
major product. The isomeric ratio of the starting material is
reflected in the product distribution, revealing that the ene
reaction of cis-1.2 is too slow and outcompeted by the
intermolecular reaction of the aryne intermediate with an
external nucleophile. The transition state for the concerted
reaction requires proper alignment of all components, and the
15573
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580

Journal of the American Chemical Society
Article
Scheme 3. Substituent Effects
cis-olefin cannot adopt the correct conformation for reaction
due to the intramolecular nature of the reaction. Thus, the
geometry of the olefin is an important factor in determining the
success of this reaction, with cis-olefins being unsuitable
reaction partners.
at C2, a so-called “mismatched” case (the site of nucleophilic
attack is not the preferred site for nucleophilic attack) (Scheme
3). The activation barrier for this process was computed to be
7.9 kcal/mol. Interestingly, reaction of substrate 1.4, which
should yield an aryne intermediate that is polarized similarly to
M2-GS, provides a mixture of the desired ene product and a
product resulting from attack of the base (Scheme 3). In an
attempt to improve the ratio in favor of the ene product the
reaction was diluted, and LDA was added via syringe pump;
however, this lead to only a modest change in the ratio and the
reaction suffered from poor conversion (Scheme 3). This result
clearly demonstrates that despite the intramolecular nature of
the ene reaction, intermolecular nucleophilic attack can become
a competitive pathway. Finally, in the case M3-GS, the aryne
ground state is similarly polarized to M2-GS, but in the
opposite direction, a so-called “matched case” (the site of
nucleophilic attack coincides with the preferred site for
nucleophilic attack). The activation barrier for this process
was computed to be 4.5 kcal/mol. Experimentally, substrate 1.5
yields a single product in good yield. Thus, one must consider
the electronic influence of substituents on the aryne when
designing a reaction; efficient reactions occur with sym-
metrical/relatively nonpolarized arynes and substrates with
“matched” reactivity.
Alongside the experimental work, we also used computa-
tional models to investigate the nature and reactivity of the
aryne intermediate. The results of DFT calculations performed
at the B3LYP/6-31G* level of theory revealed an early pseudo-
chairlike transition state that proceeded in a concerted manner;
however, bond formation was asynchronous, with C−C bond
formation being more advanced (Scheme 3).10 Recently Houk
and Garg have put forth a distortion model to predict the site
selectivity of nucleophilic additions to arynes.11,12 The presence
of a σ-electron withdrawing substituent distorts the ground
state structure of the aryne, causing nucleophilic attack to occur
at the site with a larger internal angle because less energy is
needed to distort the ground state into the transition state
structure. From the computational and experimental data, it is
apparent that the aryne ene reaction resembles a nucleophilic
attack on the aryne, because the electronic nature of the aryne
has a dramatic effect on the outcome of the reaction. In the case
of M1-GS the aryne ground state does not suffer much
distortion and the activation barrier was computed to be 4.0
kcal/mol (Scheme 3). Experimentally, the reaction of substrate
1.3, which should yield a relatively symmetrical aryne
intermediate, goes on to provide a single product in excellent
yield. In the case of M2-GS, the aryne is polarized toward the
oxygen atom such that nucleophilic attack would be preferred
Scope of the Reaction. With this knowledge, we went on
to explore the scope of the reaction (Table 3). In order to
achieve a selective deprotonation and efficiently generate the
aryne, a directing group is necessary that can precomplex the
lithium amide base.13 We were interested in exploring the
15574
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580

Journal of the American Chemical Society
Article
a
Table 3. Scope of the Intramolecular Aryne Ene Recation
a
Reaction conditions: substrate (1 equiv), LDA (1.1 equiv, 0.1 or 0.5 M in THF), THF (0.024 or 0.05 M or 0.10 M, based on substrate), rt (See
b
c
d
Supporting Information for specific examples). Isolated yields. Amine addition competes (32% of 5.2). Amine addition competes (60% of 5.3).
e
f
LDA (2.2 equiv). Amine addition competes (13% of 5.4).
directing effects of substituents on the substrate and probing
the limits of the deprotonation based approach of aryne
generation. When we remove the directing group, benzylic
deptotonation occurs and the product of [2,3]-Wittig
rearrangement results (entry 2), demonstrating that the
bromine on its own is not a sufficiently strong directing
group. In the reaction of substrate 1.7, the bromine has been
moved to the meta-position; two possible regioisomeric arynes
could be formed. It was thought that the ether oxygen and
bromine would direct in concert to selectively deprotonate the
ortho-position; however, benzylic deprotonation occurred to
give 3.3 (entry 3). When the position of the oxygen is moved,
as in substrate 1.4, the bromine and oxygen direct in concert to
give a selective deprotonation; however, in this case the yield is
low due to competing addition of external base (entry 4). In the
case of substrate 1.8, the Wittig rearrangement cannot take
place, and the desired ene reaction takes place in good yield
(entry 5). In this case, one could imagine a scenario where an
equilibrium exists with other lithiated species; however, the
elimination of the bromine is irreversible; therefore the desired
pathway is traversed. Substrates 1.9 and 1.10 demonstrate the
directing effect of two halogens and the selective elimination of
the bromine to give predominately one aryne, which undergoes
the ene reaction (entries, 6 and 7). In the case of substrate 1.10,
15575
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580

Journal of the American Chemical Society
Article
the chlorine provides a synthetic handle. The pyrazolyl group
proved to be an excellent directing group, whereas the teriary
amine was not (entries 8 and 9). Significantly, substrate 1.13,
bearing a basic tertiary amine, reacted smoothly to give the
heterocycle 2.9 in excellent yield (entry 10). We then went on
to explore the ring sizes that are accessible using this
methodology. The reaction of substrate 1.14 proceeded cleanly
to give a mixture of the desired cyclized product in 35% and a
product resulting from external attack by base in 60%,
demonstrating that cyclization to make seven-membered rings
is slow and competing nucleophilic attack can compete (entry
11). However, the reaction of substrate 1.15 proceeded
smoothly to give the cyclized product 2.11 in 90% yield
(entry 12). In the five-membered ring series, the reaction of
substrate 1.16 proceeded to give starting material and product
4.2, which arises from an ene reaction followed by further
reaction of the ene product. Under the standard reaction
conditions, none of the ene product was observed, likely
because of the good leaving group ability of the phenolate
anion and the strain released when opening the five-membered
ring. A good yield of 4.2 could be obtained by addition of 2.2
equiv of LDA at the outset of the reaction (entry 13). If an all
carbon linker is utilized, as in substrate 1.17, then a good yield
of the cyclized product can be obtained (entry 14). We were
interested in extending the intramolecular ene cyclization
strategy to include the use of hetarynes (entries 15−17).
Reaction of substrate 1.18 proceeded in a modest 50% yield
with some (13%) competing addition of the external base to
the 4-position (entry 15). This competing nucleophilic attack is
a result of the polarization of the aryne triple bond toward the
ring nitrogen, which makes the carbon at the 4-position more
electrophilic. Substrate 1.20 was selected because the fluorine
atom adjacent to the newly formed aryne would counteract the
effect of the indole nitrogen, thereby making the aryne more
symmetrical and allowing for an efficient reaction to take place.
A similar strategy was used by Garg in the synthesis of
indolactam V, where a bromine substituent was used to
improve the site selectivity of nucleophilic attack and could
later be removed.14 Reaction of substrate 1.20 proceeded to
give the ene product in quantitative yield (entry 17).
Scheme 4. Cyclization of Amides, Ureas, and Carbamates
a
Table 4. Alternate Methods of Aryne Generation
substrate
b
entry
X
Z
RM
yield (%)
1
2
3
OSO2-4-Cl-Ph
OSO2-4-Cl-Ph
Cl
Br
Br
H
1.41
1.41
1.42
iPrMgCl·LiCl
nBuLi
84
65
80
nBuLi
a
Substrate (1 equiv), RM (1.05 equiv), THF (0.1 M, based on
b
substrate), −78 °C for 3 h then rt for 18 h. Isolated yield.
Amides and carbamates are excellent directing groups for
directed ortho-lithiation, and we thought to use them to guide
selective aryne formation; however, substrates of this type did
not provide any desired product and only gave complex
mixtures. This is likely due to cyclization of the carbonyl group
onto the aryne (Scheme 4). It has been demonstrated that
benzoxazole and benzothiazole type products can be formed by
intramolecular cyclization of the carbonyl compound onto the
aryne followed by quenching with electrophiles.15
We also explored alternate methods of aryne generation and
found that protocols based on metal−halogen exchange and
deprotonation with alkyl lithiums provided the desired products
in comparable yields to deprotonation with LDA (Table 4).
The absence of nucleophilic species after the deprotonation
may prove advantageous in avoiding side reactions.
With prior knowledge that the olefin geometry is important
in the ene reaction, we explored the variation of substituents on
the olefin (Table 5). An interesting spirocyclic compound
bearing an all carbon quaternary center could be formed in
good yield (entry 1). A styrenyl olefin was tolerated (entry 2).
Substrate 1.23 bearing two similar alkyl substituents reacted
regioselectively to give 2.18 in good yield (entry 3).
Interestingly, a vinyl silane could also be employed in the
reaction (entry 3), to give an isomerized vinyl silane, which is
poised for further manipulation. In this case, 1.24 was obtained
as an inseparable mixture of E/Z isomers, but we have
demonstrated that only the Z-isomer will react; we have based
the yield of 2.19 on the amount of Z-isomer in the mixture. Up
to this point, the ene reactions unambiguously gave a single
product bearing a terminal olefin. We were interested in the
possibility of controlling oelfin geometry. We synthesized
substrate 1.25, which would be expected to give an internal
olefin following the ene reaction. Modeling the reaction
revealed two transitions states, TS1 and TS2, which were
quite close in energy (Figure 2). TS1 suffers from 1,2-allylic
strain between the two eclipsing methyl groups and would lead
to the E-olefin geometry; on the other hand, in TS2 the 1,2-
allylic strain is minimized, but it suffers from 1,3-allylic strain
and would lead to the Z-olefin geometry. The energy difference
between these competing interactions will dictate the product
distribution. Computationally there was found to be an energy
difference of 0.53 kcal/mol, corresponding to a product
distribution of 1:2.4 (E/Z). Experimentally, substrate 1.25
gives a 1:2 ratio of E/Z-isomers in favor of the Z-olefin, this
ratio could not be improved on conducting the reaction at
temperatures as low as −78 °C.
15576
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580

Journal of the American Chemical Society
Article
a
Table 5. Scope of the Intramolecular Aryne Ene Reaction
Figure 3. Model of a Stereoselective Aryne Ene Reaction.
stereocenter is to populate a chair conformation in which the
allylic substituent (methyl in the model) is placed pseudoe-
quatorially; rotation about the single bond then provides access
to either diastereomer. TS4 suffers from 1,3-allylic strain,
raising its energy 4.19 kcal/mol above TS3. An energy
difference of this magnitude should result in the observation
of only one diastereomer and the relative stereochemistry is
predicted to be trans.
These predictions were confirmed experimentally: when a
substituent was present at the allylic position, diastereoselectiv-
ities of >20:1 in favor of the trans-product were observed
(Table 6, entries 1−5). Substrate 1.31 reacted to give product
a
Table 6. Stereoselective Aryne Ene Reaction
a
Reaction conditions: substrate (1 equiv), LDA (1.1 equiv, 0.1 or 0.5
M in THF), THF (0.024 or 0.05 M or 0.10 M, based on substrate), rt
b
(See Supporting Information for specific examples). Isolated yields.
c
d
Yield based on Z-isomer. 1:2, E/Z determined by 1H NMR analysis
of the crude reaction mixture.
a
Reaction conditions: substrate (1 equiv), LDA (1.1 equiv, 0.1 or 0.5
M in THF), THF (0.024 or 0.05 M or 0.10 M, based on substrate), rt
b
(See Supporting Information for specific examples). Determined by
c
1H NMR analysis of the crude reaction mixture. Isolated yields.
2.26 bearing an all carbon quaternary center in good yield as a
1.4:1 mixture of isomers. The presence of the methyl group on
the olefin reduces the energy difference between the two
diastereomeric transition states, leading to an unselective
reaction. When the stereocenter is placed at the benzylic
position, poor selectivity is observed (entries 7 and 8), which is
not so surprising, given that it is more remote. Stereoselective
reactions of arynes are rare in the literature, and this may be
Figure 2. Model reaction to provide an internal olefin.
We were interested in the possibility of a substrate-controlled
stereoselective ene reaction. We modeled the reaction of a
substrate bearing an allylic stereocenter and DFT calculations
revealed two pseudo-chairlike transition states, TS3 and TS4,
with significantly different energies (Figure 3). The effect of the
15577
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580

Journal of the American Chemical Society
Article
due to the commonly held notion that a highly reactive
substrate is not very selective; however, this does not appear to
be the case. The origin of the high selectivity is largely a result
of the substrate’s preference for a single low-energy ground
state conformation, explaining why selectivity can be achieved
despite an early, reactant-like transition state.
Formal Synthesis of ( )-Crinine. With the knowledge
garnered from our mechanistic and substrate studies, we sought
to validate the utility of this methodology in the context of the
synthesis of a natural product. We identified crinine as a
potential target that could be accessed via an intramolecular
aryne ene reaction (Figure 4). Crinine is a member of the
Figure 5. Retrosynthesis.
using the ene reaction, so we had some confidence that this key
step would succeed. Key intermediate 1.35 can be disconnected
into the pyrrolidine 1.34 and commercially available 6-
bromopiperonal.
Figure 4. ( )-Crinine and Common Disconnection.
The first task in the synthesis was to synthesize compound
1.35 succintly, in order to test the key ene reaction (Scheme 5).
5,10b-ethanophenanthridine class of the amaryllidaceae family
of alkaloids, characterized by its unique skeleton. Members of
the amaryllidaceae family of alkaloids have a number of
interesting and varied biological activities.16 The compound
was first isolated by Wildman in 1955 from the bulbs of two
unidentified crinum species from South Africa. In a remarkable
study Wildman demonstrated the skeletal connectivity of this
Scheme 5. Forward Synthesis
alkaloid by degradation and chemical correlation studies.17
A
number of syntheses have relied on strategic disconnection to
the C3-arylpolyhydroindole substructure (Figure 4). The first
total synthesis of crinine was reported by Muxfeldt and co-
workers in 1966 .18 In addition to providing support for the
assigned structure and defining the relative configuration of
crinine, the synthesis demonstrated the strategic disconnection
to the C3-arylpolyhydroindole substructure, laying the ground-
work for many syntheses that followed. Since this time a
number of approaches to ( )-crinine have been described,19
many of them relying on accessing a C3-arylpolyhydroindole
motif. Furthermore, many elegant approaches have been
developed to access other members of the ethanophenan-
thridine class of natural products.20
The plan we devised relied on intercepting intermediate
1.38, which was used in the Whitlock19a synthesis of
( )-crinine (Figure 5). This intermediate provided a way in
which to build the C-ring onto the B-ring, which would
ultimately be created by the ene reaction. In the forward
direction, a sequence of intramolecular aldol reaction of 1.37
followed by dehydration could provide 1.38, which would
constitute a formal synthesis of crinine. Ketoaldehyde 1.37
could be traced back to 1.36 by ozonolytic cleavage.
Disconnection of the key bond to build the B-ring leads back
to compound 1.35, which is poised for the ene reaction.
Inspection of computer and hand-held models revealed that the
presence of the pyrrolidine ring in 1.35 would secure the
correct relative stereochemistry because the allyl group would
prefer to occupy a pseudoequatorial position. We have seen
previously that quaternary centers could be easily installed
The synthesis began with commercially available N-Boc-
pyrrolidinone, which underwent an aldol reaction with acetone
to give the adduct, which was used without purification and
dehydrated by formation of the mesylate and in situ elimination
to give pyrrolidinone 1.39 in a moderate, 54%, overall yield.
The oxidation state of 1.39 was adjusted by action of DIBAL
and the crude reaction mixture was directly allylated to give
compound 1.40 in 52% yield over two steps. Pyrrolidinone
1.40 was deprotected under standard conditions, and the amine
was used directly without any further purification. Reductive
15578
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580

Journal of the American Chemical Society
Article
amination of 6-bromopiperonal in the presence of NaBH-
(OAc)3 proceeded in moderate yield, likely due to the steric
hindrance associated with the secondary amine. Other
reduction procedures failed to improve the result. With 1.35
in hand, the key aryne ene reaction could be tested, and we
were pleased to see a relatively clean reaction with the
formation of the desired product 1.36, in 50% yield, as a single
diastereomer with the correct relative configuration. The
moderate yield of product 1.36 relative to some of the acyclic
congeners (see previous section) is thought to be a result of
increased strain in the transition state associated with the
formation of the bicyclic structure.
Scheme 6. Completion of Synthesis of ( )-Crinine
All that remained in order to finish the formal synthesis of
( )-crinine was the projected ozonolytic cleavage of both
olefins, followed by aldol cyclization and dehydration. We
found that ozonolysis was complicated by the electron-rich
aromatic ring and teriary amine. So we attempted to attenuate
the reactivity by using the hydrochloride salt of 1.36
(1.36·HCl). Ozonolysis of 1.36·HCl at −78 °C in DCM
followed by reductive work up led to the formation of 1.43 in
23% yield, another compound that possessed a molecular
formula corresponding to over-oxidation, in approximately
equal amounts (27%), and the rest of the mass balance was
mainly decomposition products. An ozonolysis experiment was
also attempted by using SudanRed 7B as an indicator to avoid
over-oxidation, but this gave identical results to the reaction
without any indicator. Switching the solvent to MeOH or
EtOAc did not lead to improved results, and there was
extensive decomposition of the starting material. Under several
conditions for dihydroxylation or oxidation with mCPBA, only
oxidation of the 1-substituted olefin took place, indicating that
the 1,1-disubstituted olefin was more sterically hindered than
we anticipated at the outset of the synthesis. In order to
circumvent the difficult oxidation, we recognized that
cyclization via a carbonyl ene reaction21 could lead to a change
in the conformation of the molecule, providing more
accessibility to the olefin. We screened a number of Lewis
acids such as Me2AlCl, EtAlCl2, SnCl4, TiCl4, Sc(OTf)3, MAD,
and MADPH; however, only starting material or decom-
position was observed. A report from the Overman group
demonstrated a difficult type II carbonyl ene reaction of a
ketone, using aluminum trichloride as a promoter. It is
noteworthy that this is one of the few examples where a
carbonyl ene reaction has be carried out with a tertiary amine
present in the molecule.22
6). Hydroxy ketone 1.45 could be dehydrated in a
straightforward manner to give compound 1.38, which was
previously prepared by Whitlock and co-workers in their
synthesis of ( )-crinine,19a constituting a formal synthesis of
the natural product.
CONCLUSIONS
■
The development of a general and high yielding method for the
formation of various carbo- and heterocycles based on an
intramolecular aryne ene reaction has been disclosed. The
intramolecular strategy has allowed for high levels of chemo-,
regio-, and stereoselectivity to be achieved. Deuterium labeling
experiments, carefully designed substrates, and calculations
have provided insight into the reaction mechanism. The
knowledge gained in our mechanistic and substrate studies was
put to use in a formal synthesis of the alkaloid ( )-crinine. The
synthesis showcases the use of the aryne ene reaction in a key
step where an all carbon quanternary center is set and the
relative configuration of both stereocenters is secured in a
single step, vastly increasing the complexity of the precursor.
However, the synthesis is later plagued by the unforeseen
difficulty of oxidizing the 1,1-disubstituted olefin. Although the
synthesis suffers from a low yield following the key step, it
successfully shows how the aryne ene reaction can be applied in
target oriented synthesis. We are currently trying to expand the
repertoire of annulation strategies using aryne intermediates to
form C−C bonds and further the use of arynes in stereo-
selective synthesis.
We were pleased to find that using AlCl3 in DCE resulted in
cyclization to give 1.44 in 21% yield as a single diastereomer
(Scheme 6). During the optimization of this step, high
conversions and low yields were always observed. Due to the
1
lack of observable byproducts in the H NMR spectrum, it is
likely that polymerization of the substrate takes place or there is
extensive decomposition, leading to insoluble byproducts. The
poor yields in this reaction are likely due to the basic tertiary
amine, which complexes the Lewis acid and leads to side
reactions. Although the yield of the carbonyl ene reaction was
low, we were interested in determining whether our initial
hypothesis was correct. We subjected 1.44 to dihydroxylation
conditions and found that this compound smoothly reacted to
give one product by 1H NMR analysis, in essentially
quantitative yield. The glycol could then be cleaved with
sodium periodate in moderate yield to provide compound 1.45,
whose structure was proven by X-ray crystallography,
confirming our previous stereochemical assignments (Scheme
ASSOCIATED CONTENT
* Supporting Information
Experimental procedures, spectroscopic and analytical data for
all products and starting materials, copies of NMR spectra, and
X-ray data for compound 1.45. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
mlautens@chem.utoronto.ca
Notes
■
The authors declare no competing financial interest.
15579
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580

Journal of the American Chemical Society
Article
VanderVelde, D.; Yang, S.; Brassfield, A.; Buszek, K. R. Tetrahedron
Lett. 2009, 50, 63. (c) Garr, A. N.; Luo, D.; Brown, N.; Cramer, C. J.;
Buszek, K. R.; VanderVelde, D. Org. Lett. 2010, 12, 96.
(13) Snieckus, V. Chem. Rev. 1990, 90, 879.
(14) Bronner, S. M.; Goetz, A. E.; Garg, N. K. J. Am. Chem. Soc. 2011,
133, 3832.
(15) (a) Clark, R. D.; Caroon, J. M. J. Org. Chem. 1982, 47, 2804.
(b) Stanetty, P.; Krurnpak, B. J. Org. Chem. 1996, 61, 5130.
(c) Fairhurst, R. A.; Janus, D.; Lawrence, A. Org. Lett. 2005, 7, 4697.
(16) (a) Ghosal, S.; Saini, K. S.; Razdan, S. Phytochemistry 1985, 24,
2141. (b) Kornienko, A.; Evidente, A. Chem. Rev. 2008, 108, 1982.
(c) Banwell, M. G.; Gao, N.; Schwartz, B. D.; White, L. V. Top. Curr.
Chem. 2012, 309, 163.
(17) (a) Mason, L. H.; Puschett, E. R.; Wildman, W. C. J. Am. Chem.
Soc. 1955, 77, 1253. (b) Wildman, W. C. J. Am. Chem. Soc. 1958, 80,
2567.
ACKNOWLEDGMENTS
■
We thank the Merck Frosst Center for Therapeutic Research,
the Natural Science and Engineering Research Council
(NSERC), and the University of Toronto for financial support.
D.A.C. thanks NSERC for a CGS D fellowship. D.D. thanks
NSERC for an undergraduate student research award.
REFERENCES
■
(1) Reviews: (a) Dyke, A. M.; Hester, A. J.; Lloyd-Jones, G. C.
Synthesis 2006, 4093. (b) Chen, Y.; Larock, R. C. In Modern Arylation
Methods; Akermann, L., Ed.; WILEY-VCH: Weinheim, Germany,
2009; pp 401−473. (c) Sanz, R. Org. Prep. Proced. Int. 2008, 40, 215.
(d) Pellissier, H.; Santelli, M. Tetrahedron 2003, 59, 701.
(e) Hoffmann, R. W. Dehydrobenzene and Cycloalkynes; Academic
Press: New York, 1967. (f) Gilchrist, T. L. In Science of Synthesis; Hopf,
H., Ed.; Georg Thieme Verlag KG: Stuttgart, Germany, 2011; pp 151−
203. (g) Tadross, P. M.; Stoltz, B. M. Chem. Rev. 2012, 112, 3550.
(h) Kitamura, T. Aust. J. Chem. 2010, 63, 987. (i) Gampe, C. M.;
Carreira, E. M. Angew. Chem., Int. Ed. 2012, 51, 3766.
(2) (a) Himeshima, Y.; Sonoda, T.; Kobayashi, H. Chem. Lett. 1983,
1211. (b) Uchiyama, M.; Miyoshi, T.; Kajihara, Y.; Sakamoto, T.;
Otani, Y.; Ohwada, T.; Kondo, Y. J. Am. Chem. Soc. 2002, 124, 8514.
(c) Lin, W.; Chen, L.; Knochel, P. Tetrahedron 2007, 63, 2787.
(d) Sapountzis, I.; Lin, W.; Fischer, M.; Knochel, P. Angew. Chem., Int.
Ed. 2004, 43, 4364.
(18) Muxfeldt, H.; Schneider, R. S.; Mooberry, J. B. J. Am. Chem. Soc.
1966, 88, 3670.
(19) (a) Whitlock, H. W.; Smith, G. L. J. Am. Chem. Soc. 1967, 89,
3600. (b) Overman, L. E.; Jacobsen, E. J. Tetrahedron Lett. 1982, 27,
2741. (c) Overman, L. E.; Sugai, S. Helv. Chim. Acta 1985, 68, 745.
(d) Martin, S. F.; Campbell, C. L. Tetrahedron Lett. 1987, 28, 503.
(e) Martin, S. F.; Campbell, C. L. J. Org. Chem. 1988, 53, 3184.
(f) Pearson, W. H.; Lovering, F. E. Tetrahedron Lett. 1994, 35, 9173.
(g) Pearson, W. H.; Lovering, F. E. J. Org. Chem. 1998, 63, 3607.
(h) Bohno, M.; Imase, H.; Chida, N. Chem. Commun. 2004, 1086.
(i) Bohno, M.; Sugie, K.; Imase, H.; Yusof, Y. B.; Oishi, T.; Chida, N.
Tetrahedron 2007, 63, 6977. (j) Tam, N. T.; Chang, J.; Jung, E.-J.;
Cho, C.-G. J. Org. Chem. 2008, 73, 6258. (k) Tam, N. T.; Cho, C.-G.
Org. Lett. 2008, 10, 601. (l) Liu, J.-D.; Wang, S.-H.; Zhang, F.-M.; Tu,
Y.-Q.; Zhang, Y.-Q. Synlett 2009, 3040. (m) Guillou, C.; Bru, C.
Tetrahedron 2006, 62, 9043. (n) Pandey, G.; Gupta, N. R.; Gadre, S. R.
Eur. J. Org. Chem. 2011, 740.
(20) For recent reports aimed at the synthesis of ethanophenan-
thridine alkaloids, see: (a) Yang, L.; Chen, W.; Wang, X.; Pan, Z.;
Zhou, M.; Yang, X. Synlett 2011, 207. (b) Findlay, A. D.; Banwell, M.
G. Org. Lett. 2009, 11, 3160. (c) Roe, C.; Stephenson, G. R. Org. Lett.
2008, 10, 189. (d) Gao, S.; Tu, Y. Q.; Song, Z.; Wang, A.; Fan, X.;
Jiang, Y. J. Org. Chem. 2005, 70, 6523. (e) Song, Z. L.; Wang, B. M.;
Tu, Y. Q.; Fan, C. A.; Zhang, S. Y. Org. Lett. 2003, 5, 2319. (f) Banwell,
M. G.; Harvey, J. E.; Jolliffe, K. A. J. Chem. Soc., Perkin Trans. 1 2001,
2002. (g) Ley, S. V.; Schucht, O.; Thomas, A. W.; Murray, P. J. J.
Chem. Soc., Perkin Trans. 1 1999, 1251. (h) Schkeryantz, J. M.;
Pearson, W. H. Tetrahedron 1996, 52, 3107. (i) Pearson, W. H.;
Lovering, F. E. J. Am. Chem. Soc. 1995, 117, 12336. (j) Grigg, R.;
Santhakumar, V.; Sridharan, V.; Thornton-Pett, M.; Bridge, A. W.
Tetrahedron 1993, 49, 5177. (k) Burk, R. M.; Overman, L. E.
Heterocycles 1993, 35, 205.
(3) (a) Arnett, E. M. J. Org. Chem. 1960, 25, 324. (b) Simmons, H. E.
J. Am. Chem. Soc. 1961, 83, 1657. (c) Wittig, G.; Durr, H. Justus Liebigs
̈
Ann. Chem. 1964, 672, 55. (d) Kampmeier, J. A.; Rubin, A. B.
Tetrahedron Lett. 1966, 2853. (e) Nakayama, J.; Yoshimura, K.
Tetrahedron Lett. 1994, 35, 2709. (f) Aly, A. A.; Mohamed, N. K.;
Hassan, A. A.; Mourad, A.-F. E. Tetrahedron 1999, 55, 1111.
(4) (a) Aly, A. A.; Shaker, R. M. Tetrahedron Lett. 2005, 46, 2679.
(b) Jayanth, T.; Jeganmohan, M.; Cheng, M.; Chu, S.; Cheng, C. J. Am.
Chem. Soc. 2006, 128, 2232.
(5) (a) Chen, C.-L.; Sparks, S. M.; Martin, S. F. J. Am. Chem. Soc.
2006, 128, 13696. (b) Per
Tetrahedron Lett. 1990, 31, 2331. (c) Gonzal
́
ez Meiras
́
, D.; Guitian
́
, E.; Castedo, L.
́
ez, C.; Per
́
ez, D.; Guitian
́
,
E.; Castedo, L. J. Org. Chem. 1995, 60, 6318. (d) Buszek, K. R.
Tetrahedron Lett. 1995, 36, 9125. (e) Hayes, M. E.; Shinokubo, H.;
Danheiser, R. L. Org. Lett. 2005, 7, 3917.
(6) Our preliminary Communication: Candito, D. A.; Panteleev, J.;
Lautens, M. J. Am. Chem. Soc. 2011, 133, 14200.
(7) Nakai, T.; Mikami, K. Chem. Rev. 1986, 86, 885.
(8) (a) Huisgen, R.; Sauer, J. Ber. Dtsch. Chem. Ges. 1959, 72, 192.
(b) Zoltewicz, J.; Bunnett, J. J. Am. Chem. Soc. 1965, 87, 2640.
(c) Dunn, G.; Krueger, P.; Rodewald, W. Can. J. Chem. 1961, 39, 180.
(9) (a) Crews, P.; Beard, J. J. Org. Chem. 1973, 38, 522. (a) Friedman,
L.; Osiewicz, R. J.; Rabideau, P. W. Tetrahedron Lett. 1968, 5735.
(b) Wasserman, H. H.; Solodar, A. J.; Keller, L. S. Tetrahedron Lett.
1968, 5597. (c) Wasserman, H. H.; Keller, L. S. Tetrahedron Lett.
1974, 4355. (d) Tabushi, I.; Okazaki, K.; Oda, R. Tetrahedron 1969,
25, 4401. (e) Ahlgren, G.; Akermark, B. Tetrahedron Lett. 1970, 3047.
(f) Garsky, V.; Koster, D. F.; Arnold, R. T. J. Am. Chem. Soc. 1974, 96,
4207. (g) Jayanth, T.; Jeganmohan, M.; Cheng, M.; Chu, S.; Cheng, C.
J. Am. Chem. Soc. 2006, 128, 2232.
(21) Clarke, M. L.; France, M. B. Tetrahedron 2008, 64, 9003.
(22) Overman, L. E.; Lesuisse, D. Tetrahedron Lett. 1985, 26, 4167.
(10) Calculations were performed using Spartan ’08 V1.2.0.
Geometry optimizations were carried out at B3LYP/6-31G* level of
theory. Frequency calculations were used to characterize stationary
points as minima or transition states. See Supporting Information.
(11) (a) Cheong, P. H.-Y.; Paton, R. S.; Bronner, S. M.; Im, G-Y. J.;
Garg, N. K.; Houk, K. N. J. Am. Chem. Soc. 2010, 132, 1267. (b) Im,
G.-Y. J; Bronner, S. M.; Goetz, A. E.; Paton, R. S.; Cheong, P. H.-Y.;
Houk, K. N.; Garg, N. K. J. Am. Chem. Soc. 2010, 132, 17933.
(c) Goetz, A. E.; Bronner, S. M.; Cisneros, J. D.; Melamed, J. M.;
Paton, R. S.; Houk, K. N.; Garg, N. K. Angew. Chem., Int. Ed. 2012, 51,
2758.
(12) (a) Buszek, K. R.; Luo, D.; Kondrashov, M.; Brown, N.;
VanderVelde, D. Org. Lett. 2007, 9, 4135. (b) Brown, N.; Lou, D.;
15580
dx.doi.org/10.1021/ja306881u | J. Am. Chem. Soc. 2012, 134, 15572−15580