Y. Zhang et al. / Tetrahedron 69 (2013) 11080e11083
11081
For a straightforward reaction, it is ideal to get the pure product
from reaction mixture directly even without recrystallzation.17
Herein, the convenience of this protocol was demonstrated in
Fig. 2, where the reaction (entry 13) started as a reddish clear so-
lution of aldehyde, amide, and BF3 etherate in acetone and ended as
a white gel. The pure solid product 1a could be collected by simple
filtration.
Scheme 1. Catalyzing the reaction with Lewis acid.
2. Results and discussion
The initial study was to screen different kinds of Lewis acids
and solvents to identify the most reactive catalyst and medium
for the condensation and the results were summarized in Table 1
where p-fluorobenzaldehyde and benzamide were employed as
standard substrate. To achieve the mild condition, the reaction
was locked at room temperature with 10 mol % catalyst. Lewis
acids, such as Ti(OiPr)4, Mg(ClO4)2, Sc(OTf)3, Cu(OTf)2, CuBr, and
TiCl4 did not achieve any positive result (entries 1e6). Widely
applied AlCl3 (entry 7) gave only 25% yield. To our delight, BF3
etherate, a well-known14 and cheap15 Lewis acid, increased the
yield significantly to 47% in diethyl ether (entry 8). The following
solvent optimization revealed that the solubility of both sub-
strate and product play an important role in this process. It was
found when the substrate was well soluble and the product could
precipitate at low concentration, the reactivity, the ease of
workup by filtration, and the high purity of product could meet
together. Thus the toluene, DCM, THF, and ethyl acetate (entries
9e12) failed to obtain homogenous solution at the first stage, and
in turn diminished the yield and purity of final product. Mean-
while highly polar solvents, such as DMF and EtOH (entries 14 &
15) also failed to enhance the reaction due to their perfect sol-
ubility for both substrate and product. To our surprise, acetone,
a conventional aldol donor and a relatively green and healthy
solvent for room temperature reaction,16 boosted the yield to 82%
(entry 13).
Fig. 2. The product 1a was achieved as white gel.
Encouraged by these results, we explored a variety of alde-
hydes and amides with BF3 etherate as the catalyst (Table 2). The
aromatic aldehyde bearing electron-withdrawing group gave
faster conversion than 3-methoxysalicylaldehyde (1a vs 1h). The
aromatic aldehydes containing bromide were favored due to the
enhanced precipitation (1def). To our delight, several functional
groups, such as phenol and thiophenyl were well compatible
without significant impact on the yield (1hej). The electronic
effect on aromatic amide was also investigated and it was found
that both electron withdrawing (2aec) and electron donating
groups (3aec, 4a,b) benefit the reactivity. The amides, such as
cinnamide, acrylamide, phenylacetamide, and thienylacetamide
gave corresponding bisamide (5a,b, 6, 7aee, 8aee) as precipitate
with high yield. Meanwhile, the fully aliphatic amides reacted
with a variety of aldehydes, offering bisamides (9aee, 10a,b) in
the same manner.
Table 1
The screening for optimum catalyst and solventa
The catalytic efficiency of BF3 etherate was measured with the
reaction toward 7b employing catalyst loading from 10 mol % to
0.5 mol % at room temperature. It was clear that with extended
reaction time, 0.5 mol % catalyst worked well almost without any
compromise of productivity (Table 3).
To explore the scalability of this conversion, the preparation of
1m was scaled up to 1 mol with 1 mol % BF3 etherate as catalyst. The
reaction finished in 12 h in the yield of 82% and the product was
pure as well as that from the routine scale.
Entry
Catalyst
Solvent
Yieldb
1
2
3
4
5
6
7
8
Mg(ClO4)2
Ti(OiPr)4
Sc(OTf)3
Cu(OTf)2
CuBr
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Toluene
DCM
THF
N. D.
5%
N. D.
5%
A plausible mechanism for this conversion was illustrated in
Scheme 2. The cycle commences from the coordination of al-
dehyde to the BF3 and the activated carbonyl is attacked by the
amide to form the hemi aminal intermediate A in which the
hydroxyl is removed with the help of Lewis acid and gem-amide
group. In turn, the enamide ion pair intermediate B may react
with the second amide directly to give the bisamide in-
termediate C and yield the final bisamide after dehydration. On
the other hand the enamide ion pair intermediate B may de-
hydrate prior to the nucleophilic addition of the second amide
and gave the overall same product through intermediate D. In
the presence of hydrogen bonding acceptor and lone electron
pair donor, such as acetone and diethyl ether, the reaction be-
tween water and BF3 to form fluoroboric acid can be inhibited,
which ensures the BF3 enter the next catalytic cycle.
10%
N. D.
25%
47%
62%
53%
57%
48%
82%
25%
31%
TiCl4
AlCl3
BF3*OEt2
BF3*OEt2
BF3*OEt2
BF3*OEt2
BF3*OEt2
BF3$*OEt2
BF3*OEt2
BF3*OEt2
9
10
11
12
13
14
15
EA
Acetone
DMF
EtOH
a
4-Fluorbenzaldehyde (1.0 mmol), benzamide (2.0 mmol), 0.2 M for aldehyde in
solvent. Typical catalyst loading is 10 mol %. The reaction was carried out at room
temperature for 120 min.
b
Isolated yield by filtration for entries 7e15 and NMR yield for entries 2 and 4 and
in turn diminished the yield and purity of final product. Meanwhile highly polar
solvents, such as DMF and EtOH (entries 14 & 15) also failed to enhance the reaction
due to their perfect solubility for both substrate and product. To our surprise, ace-
tone, a conventional aldol donor and a relatively green and healthy solvent for room
temperature reaction,16 boosted the yield to 82% (entry 13).
As a demonstration, CB2 receptor inverse agonists 11 and 12
were synthesized with BF3 etherate as catalyst in ether, a solvent
fine-tuned for the product (Scheme 3). The reaction gave rise to the
target as precipitate in good to excellent yield.