3102
L. O. Davis et al. / Tetrahedron Letters 55 (2014) 3100–3103
Acids such as phosphoric acid and trifluoroacetic acid were
enone 4d likely results from competing side reactions between
hydrolysis products and aldehyde 2d (Table 2, entry 3).
Less activated, branched aliphatic aldehydes proceeded with
moderate yield in the addition reaction (Table 2, entries 4 and 5).
unable to promote the reaction more than the thermal reaction
(compare Table 1, entries 2–3 to entry 1). In addition, the phospho-
ramidic acid diphenyl ester was found to be an unsuitable catalyst
for this reaction compared to diethyl phosphate (Table 1, entries 4
and 5).22 These results not only highlight the importance of diethyl
phosphate, but also demonstrate the unique function of the phos-
phoric acid catalyst.
a-Unsubstituted aldehydes (2g–2j) are a particularly challenging
class of substrates because they contain enolizable hydrogens
which can lead to the homodimerization of aldehydes. We were
pleased to find that
a-unsubstituted branched aldehydes (2g and
Interestingly, using 0.1 equiv of diethyl phosphate gave a mod-
erate yield (Table 1, entry 6),23 and a full equivalent did not
increase the yield of the enone (Table 1, entry 7). Lastly, several
solvents were screened. Toluene and acetonitrile gave inferior
reaction yields compared to those in dichloromethane (Table 1,
entries 8 and 9). Notably, during the course of the optimization
studies, suspected reaction intermediates were not detected.
We attempted the synthesis of enone 4a in a one-pot reaction
from benzylamine (5) and cyclohexanone (6), presumably through
the formation of imine 1 in situ. Compounds 5 and 6 were allowed
to stir in solution between 0 and 4 h before the acid catalyst 3 and
the aldehyde 2a were added to the reaction flask (see Supplemen-
tary data). We were pleased to find that the reaction product was
isolated in moderate yield (up to 45%), but the optimized reaction
conditions for the one-pot reaction resulted in a significantly lower
yield of 4a compared to the optimized conditions reported above
(see Table 1, entry 5). Upon further investigation, we discovered
the imine was reacting in the presence of diethyl phosphate and
aldehyde 2a to give hydrolysis products and a newly formed
aldimine 7 (Scheme 3).24 Despite our attempt to limit the presence
of water through the use of activated molecular sieves or perform-
ing the reaction under nitrogen, some hydrolysis inevitably
occurred.
The scope of the addition reaction was investigated with re-
spect to the aldehyde-component using the optimized reaction
conditions (see Table 1, entry 5). Aldehydes bearing aromatic
groups proceeded with good yield (Table 2, entries 1–3). Aldehyde
2d, bearing an electron-withdrawing group (NO2), is more
activated than the other aromatic aldehydes employed and would
therefore be expected to result in a higher yield of enone when
coupled with 1. As observed during the optimization of the
reaction conditions with aldehyde 2a, the hydrolysis of imine 1
allows for multiple reaction pathways. Considering this possibility
and the increased reactivity of aldehyde 2d, the lower yield of
2h) proceeded with moderate yield (Table 2, entries 6 and 7).
Although one might anticipate homodimerization, these products
were not detected. The coupling of hexanal (2i), a straight-chained
aliphatic aldehyde, with 1 resulted in a comparable yield to the
branched substrates (Table 2, entry 8), however a significant
decrease in yield was noted with the substrate butanal (2j) (Table 2,
entry 9). Degradation of 2j was observed,25 likely leading to the
low-yielding reaction.
The formation of enones 4a-j can be explained by two different
mechanisms: (1) a multi-step aldol-like reaction (Scheme 4a), and
(2)
a concerted, aza-ene-type mechanism (Scheme 4b). The
literature supports the possibility for both mechanisms. Babler
and co-workers suggested a step-wise mechanism in their report
of a cobalt(II) chloride catalyzed addition reaction between an
N-tert-alkyl imine derivative and several aldehydes, while Terada’s
report of
a phosphoric acid catalyzed addition reaction of
enecarbamates to imines and aldehydes suggests a concerted man-
ner.26,27 Attempts to elucidate the correct reaction mechanism
using 1H NMR titrations were not productive. The elimination step
of the reaction has also been further investigated. Our presumption
is that under truly anhydrous conditions, the elimination step
occurs before hydrolysis. A solution of b-hydroxy alcohol 1128 in
dichloromethane was stirred under a variety of reaction conditions
(Table 3). Interestingly, the aldol-condensation product 4a was not
observed under either acidic or basic conditions (Table 3, entries 1
and 2), and instead starting material was recovered in both cases.
Only when both compounds 3 and 5 were present was enone 4a
isolated (Table 3, entry 3), suggesting that dehydration occurs
through the imine; that is, elimination precedes hydrolysis. We
are currently investigating additional methods to rigorously
exclude water from the reaction mixture in an attempt to prevent
hydrolysis and isolate the imine.
In conclusion, we have reported an addition reaction between
imines and aldehydes through
a phosphoric acid catalyzed
H
step-wise
addition
Pg
3
O
NH
R2
R3
H
a
tautomerization
R1
Pg
N
Pg
H
8
2
N
OH
R3
R2
R1
R1
R2
1
Pg
H
concerted
addition
O
N
3
9
R2
R3
H
R1
b
tautomerization
8
2
elimination
O
Pg
H
N
hydrolysis
R1
R3
R1
R3
R2
R2
10
4
Scheme 4. Proposed potential mechanisms for the formation of enone 4 through either (a) a step-wise addition, or (b) a concerted addition.