2
S. Parisotto et al. / Tetrahedron Letters xxx (2015) xxx–xxx
À
OEt
El
which is an equimolar mixture of isoPr
K
NH, n-BuLi and tert-BuO -
2
LIC-KOR, THF
95°C, El-X
+
and is a weaker base in respect to LIC-KOR (entry 11). In this
−
OEt
OEt
case no product formation was observed. Finally we evaluated
the conditions indicated in entry 9 as the best found. After the
addition of the nitrone, the reaction was maintained at À95 °C
for 0.5 h, then the temperature was allowed to rise to À20 °C for
Scheme 1. General conditions for the synthesis of functionalised alkoxy-1,3-dienes
from ,b-unsaturated acetals in the presence of Li–K mixed superbases.
a
3
h. At this temperature the yield was slightly improved, maybe
was each time recovered. Diverse experimental conditions such as
a different ratio of LIC-KOR equiv in respect to the nitrone and
acetal, reaction temperatures and different time did not afford
because of the reduction of the degradation phenomena. Once
set the optimal conditions, the scope of the reaction was evaluated.
To this purpose we prepared several a,N-diarylnitrones by the acid
any addition product, leaving the nitrone unreacted. As
consequence, we turned our attention to the more electrophilic
a
catalysed condensation of the suitable substituted benzaldehyde
with N-phenylhydroxylamine. Nitrones were obtained with yields
comparable to those reported in the literature (see Supporting
a,N-diphenylnitrone 2a (Fig. 2).
4
5
We were delighted to observe that under the same experimen-
information). The results of the reaction between crotonaldehyde
diethyl acetal (3) with -aryl-N-phenylnitrones 2a–2h are shown
tal conditions previously used for nitrone 1, the reaction was com-
pleted in 2 h. The TLC monitoring of the reaction progress indicated
the disappearance of the starting nitrone. The characterisation of
the product evidenced the presence of a quaternary carbon at
a
in Table 2. As expected the presence of substituted aromatic rings
linked to the azomethyne carbon of the nitrone greatly influenced
its reactivity, especially in terms of reaction times. Para-electron-
1
3
1
2
62 ppm in the C NMR spectrum and a molecular weight of
77 AMU. The spectroscopic characterisation was coherent with
donating groups on the
the positive charge in the nitrone, thus reducing its electrophilicity
and reactivity. When -(4-N,N-dimethylaminophenyl)-N-phenyl-
nitrone 2g and -(4-methoxyphenyl)-N-phenylnitrone 2h are used
a-phenyl ring are expected to stabilise
the formation of the imine 4a represented in Scheme 2. We
hypothesised a cascade process where the addition of the nitrone
a
a
2
a to the alkoxydienyl anion is followed by the elimination of a
hydroxyl species with the formation of an imine by a E1cb process
Scheme 2). This could be reasonably promoted by LIC-KOR base on
(entries 7 and 8), only traces of the desired product are observed,
whereas the presence of a weaker electrondonating substituent
such as methyl (entry 2, nitrone 2b) affords the imine 4b in 6 h
with a 25% yield. Imine 4c and 4d are obtained in shorter times
(2.5 and 2 h, respectively, 25% and 27% yield) when 4-bromo (2c)
and 4-chloro (2d) derivatives are used (entries 3 and 4). On the
contrary, when electron-withdrawing substituents are present,
the reaction is favoured. This is confirmed by the fact that when
the halogen is in meta position, the (Z,2E)-2-ethoxy-N-phenyl-1-
(3-chlorophenyl)-penta-2,4-dien-1-imine (4e) is recovered in 2 h.
Finally an heterocyclic derivative has been prepared, but the
presence of the electron-poor pyridyl ring does not induce any
activation on the nitrone, and the corresponding imine 4f has been
obtained with a low yield (entry 6).
As can be observed, the domino process here described allows
the unsaturated imine to be obtained in a complete stereoselective
manner, where the C–C double bond has an E configuration
whereas the imine group shows a Z configuration. The geometry
of the two double bonds has been determined by a NOESY
experiment carried out on the derivative 4f, where the phenyl
and pyridine protons have different chemical shifts (6.8–7.5 ppm
and two doublets at 8.0 and 8.75, respectively). In the NOESY of
(Z,2E)-2-ethoxy-N-phenyl-(1-pyridin-4-yl)-penta-2,4-dien-1-imine
4f (reported in the Supporting information) we can observe two
correlation spots, the first clearly indicates a spatial interaction
between the methylenic group of the ethoxy moiety (3.83 ppm)
and the b-proton to the imine group (5.50 ppm), thus indicating
the E geometry of the C–C double bond (see Fig. 3, left). The second
(
the intermediate A which appears to be unstable in superbasic
medium, as witnessed by the formation of the imine 4a instead
of the expected corresponding N-hydroxylamine. To our
knowledge this type of reactivity of
a,N-diarylnitrones, which is
ascribable to the presence of the LIC-KOR base, has never been
reported. Moreover, it has to be noticed that this domino process
allows a highly conjugate imine to be obtained.
The reaction of crotonaldehyde diethyl acetal (3) with
,N-diphenylnitrone 2a was selected as model reaction in order
a
to optimise the process. This was done by evaluating the effect of
the base, both as type and number of equivalents, reaction temper-
ature and time, respectively. The results are reported in Table 1.
First of all, the role of temperature was evaluated. Whereas the
first part of the reaction was always conducted at À95 °C, attention
was paid to the temperature and reaction times after the nitrone
addition. When the reaction was carried out at À95 °C the unre-
acted nitrone was recovered at the end of the reaction (entry 1),
whereas when the temperature was increased to rt for 2.5 h a
1
6% yield (entry 2) was obtained. The yield dropped to 0% carrying
out the reaction at rt for 72 h, probably due to the decomposition
of the reagents and/or products (entry 4). We slightly changed
the temperature after the nitrone addition and observed that up
to À30 °C the rate of the process was negligible and the product
was recovered only in traces (entry 6); we obtained better results
increasing the reaction temperature at 0 °C, in this case the product
was recovered with 24% yield (entry 5). Then we evaluated the
influence of the base equivalents and, according to the mechanism
proposed in Scheme 2, we considered 3 equiv of base (entries 7 and
shows a correlation between the
c-proton in respect to the C–N
double bond (5.95 ppm) and the phenyl protons (7.00 ppm), con-
sistent with a Z configuration, (see Fig. 3, right).
8
). We obtained 24% and 26% yield, respectively. When we used a
At this point the imines 4a–4h were hydrolysed under acidic
conditions, in order to regenerate the carbonyl functionality
according to an umpolung strategy. The hydrolyses were accom-
greater amount of base, 3.2 and 4 equiv on the contrary, the yields
decreased to 12% and 0% probably because of degradation pro-
cesses (entries 10 and 3). Finally we analysed the influence of the
base and we carried out the reaction using the LIDAKOR base,
Ò
plished at rt in CHCl3 overnight using 10 meq of Amberlyst 15
1
as proton source. The H NMR analysis of the product suggested
O
t-Bu
O
Ph
H
O
R
R'
H
O
R
R'
H
N
N
N
N
Ph
H
Ph
1
2a
Figure 1. General structure of nitrones.
Figure 2. N-tert-Butyl-a-phenylnitrone 1 and a,N-diphenylnitrone 2a.