N. Jeedimalla et al. / Tetrahedron Letters xxx (2013) xxx–xxx
3
OMe
H
N
OMe
5 (1.0 equiv.)
H
O
H
O
10 mins,
pentanol, 120 o
2 hours
C
NH2
N
Cl
MeO
2a
OMe
7b
O
(31% yield)
then
3a
(1.0 equiv.)
30 mins
Cl
(1.0 equiv.)
Cl
MeO
2b
3b
MeO
OMe
OMe
not observed
4a
(1.0 equiv.)
OMe
Scheme 2. Control experiment highlighting the dual role of aniline 3 in the cascade transformation producing 4-aza-podophyllotoxin 2a.
O
H
N
O
method A:
CHO
typical one pot
NH2
H
O
2-pentanol, 120 oC
5
R1
OH
O
2c-g
R1
3c-f
, R1 = 3-SMe
4a-c
method B:
R2
NH2
sequential addition using
R2
3c
3d
3e
4a, R2 = H
4b
4c
the sacrificial aniline 3b
1
, R = 3,4-C4H8
2
, R = 3,4-OCH2O-
Cl
, R1 = 3-OH
, R2 = 3,4,5-OMe
3f, R1 = 3-OCHF2
H
N
H
N
H
N
H
N
H
F
MeS
N
O
HO
MeS
H
O
H
H
O
O
H
H
O
O
F
O
O
O
O
O
2d
2e
2f
2g
MeO
2c
O
O
O
OMe
O
O
MeO
O
method A: 12% yield
method B: 39% yield
method A: 18% yield
method B: 49% yield
method A: <2% yield
method B: 7% yield
method A: 15% yield
method B: 25% yield
method A: 7% yield
method B: 21% yield
Scheme 3. Comparative synthesis of novel 4-aza-podophyllotoxins 2c–2g between the typical and the new one-pot protocol references.
(Knoevenagel condensation) before being regenerated and con-
densed to form the product B ring (proposed mechanism in
Fig. 2). To test this hypothesis, a competitive reactivity study of
anilines was designed (Scheme 2). A sacrificial electron deficient
p-chloroaniline 3b was condensed with the same benzaldehyde
derivative 4a generating imine 7b over the course of 2 hours at
120 °C. In the same reaction vessel was then added successively
tetronic acid 5 and aniline 3a in equimolar ratio which provided
to our delight 4-aza-2,3-dehydropodophyllotoxin 2a exclusively
in 31% yield (the other possible product 2b was not detected in
the crude reaction mixture by 1H NMR or mass spectrometry anal-
yses). This result was encouraging and provides a strong support
for the dual role played by the aniline substrate 3 in the cascade
reaction.7 Incorporating this information, we therefore propose
the mechanism outlined in Figure 2. The first condensation most
likely occurs between aniline 3 and benzaldehyde 4 to deliver acti-
vated imine 7 which further endures the addition of tetronic acid 5.
Through the addition of 5, aniline 3 is then regenerated and may
then condense either on the b-position of the Michael acceptor 9
or on the carbonyl (1,4 vs 1,2 addition). To better understand this
information (even small ee) may be retained in the final product
2a, which was not observed.
Sacrificial aniline 3b in the tandem cascade reaction
After further confirmation of our mechanistic hypothesis for the
cascade reaction, we decided to examine the effect of the sacrificial
electron deficient aniline 3b on the cascade transformation scope
for several other substrates. We proposed that the successful for-
mation of the Knoevenagel intermediate 9 may also result in the
overall reaction improvement. After considerable experimental ef-
forts to optimize the stepwise cascade process (Scheme 3), we
were pleased that the more reluctant anilines 3c–f tested in the
reaction were successively condensed to yield the new 4-aza-2,3-
dehydropodophyllotoxin derivatives 2c–g in 39%, 49%, 25%, 21%
and 7% yields respectively (method B).6 For the comparative study,
all the new 4-aza-2,3-dehydropodophyllotoxins 2c–g were iso-
lated in average with threefold higher yields than using the typical
one pot procedure (isolated yields after the same purification by
method A versus method B) or a straightforward stepwise ap-
proach without sacrificial aniline 3b.9,10 Most of the compounds
2c–g synthesized via the method B were simply recrystallized from
ethanol for optimum purity affording the recovered ‘sacrificial’ ani-
line 3b as major by-product from the mother liquors.6
The cascade transformation discussed herein, highlights a novel
stepwise condensation of anilines with aldehydes and tetronic acid
resulting in 4-aza-2,3-dehydropodophyllotoxins. Several control
experiments shed light on the proposed mechanism and revealed
the order of reactivity between the three components. Our mecha-
nistic investigation resulted in the design of a new process incor-
porating an electron deficient, ‘sacrificial aniline’, enabling a
simple and versatile protocol for the synthesis of problematic
aza-analogues of podophyllotoxin (1). Using this strategy, a series
step we repeated experiments reported by Shi,4l using
a catalyst. In this reaction -proline should react in the similar fash-
L-proline as
L
ion as the aniline 3 by a LUMO lowering effect to facilitate the
nucleophilic addition on a chiral Michael acceptor (similar to 9).
For reasons that we do not explain yet, the proline catalysis pro-
ceeded poorly for our substrates leading to low isolated yields.8
As a result and in total agreement with Shi’s report, product 2a
was isolated in a racemic form (no enantiomeric excess (ee) was
detected)8 persuading us to propose that the condensation occurs
at the carbonyl carbon-centre to form 10, leading after elimination
to either quinolinium methide 11 or iminium 12 which do not car-
ry any chiral information towards the final product 2. If the con-
densation arises at the b-position, we believe that some chiral