enolate-driven domino reactivity pattern biased by the
mesomeric form 3 could take place if this N-cyclization
could be blocked. Under these conditions, the O-cyclization
reaction of the enamine 2 would provide the cyclic oxonium
ion 7, which could be conveniently transformed into differ-
ent products. This new scenario would enable the develop-
ment of new domino processes and therefore, it would in-
crease the synthetic power of these C7 platforms. The chemi-
cal accomplishment of this goal requires: 1) disfavoring the
N-cyclization; 2) favoring the shift of the electronic density
from the nitrogen to the oxygen atom; 3) stabilizing the
cyclic oxonium ion 7 by a redistribution of the electronic de-
ficiency through the canonical form 8 and 4) a polar and
protic medium to favor the separation of charges produced
going from 2 (neutral) to 3 (zwitterionic) and to assist in the
acid–base quenching of the O-cyclization reaction. Whereas
secondary amines could meet the first three requirements
through the interconversion between their corresponding
iminium and enamine forms,[10] alcohols would guarantee
the fourth condition. In this communication we report the
feasibility of this conceptual approach and we show how it
can be implemented for the diversity-oriented synthesis of
polysubstituted cyclohexadienones and multivalent aromatic
scaffolds, which possess a survey of valuable functionalities
decorating the ring.
work studying the reaction of diyne 1a (1 equiv) and pyrro-
lidine (9a; 1 equiv) in dichloromethane. After several
assays, it was found that the reaction at room temperature
for 16 h afforded the mixture of cyclohexadienones deriva-
tives 10aa and 11aa although in low yield (Table 1, entry 1).
When the amount of pyrrolidine was doubled (to adjust the
stoichiometry to the formation of 11aa), the reaction exclu-
sively generated the compound 11aa in 67% yield (Table 1,
entry 2). The structures of these two products,[12] which were
unexpected, constituted experimental evidence of a reactivi-
ty shift in the diynic platform, and they established the
chemical outcome for the novel domino reaction pathway
elicited from these platforms. According to our initial hy-
pothesis, alcohols would be beneficial solvents for this
domino reaction. It was gratifying to find that the use of
MeOH or EtOH as a solvent allowed the diamine derivative
11aa to be obtained in nearly quantitative yield (Table 1,
entries 3–4). On the other hand, the reaction of diyne 1a
with dibenzylamine (9b) was shown to be solvent depen-
dent. Whereas the reaction in EtOH generated the mixture
of monoamines 10ab (5%) and 12ab (5%) together with
the diamine 11ab (86%, Table 1, entry 5), the reaction in
MeOH afforded the monoamine 10ab (39%) and diamime
11ab (53%, Table 1, entry 6).
With these results at hand, it soon became evident that
strong nucleophilic amines and/or less nucleophilic alcoholic
solvents favored the formation of diamine derivatives 11,
whereas low nucleophilic amines and a stronger nucleophilic
alcoholic solvent favored the formation of monoamine de-
rivatives 10.
The hypothesis was assessed using the reaction of diyne
1a with different secondary amines under different reaction
conditions. Table 1 summarizes the main experimental re-
sults,[11] using two representative secondary amines: pyrroli-
dine (9a), as a representative example of cyclic and highly
nucleophilic secondary amines, and dibenzylamine (9b) as a
representative example of acyclic, sterically demanding sec-
ondary amines with reduced nucleophilicity. We began this
When dibenzylamine was used, the presence of product
12ab with an ethoxy group, revealed the participation of the
solvent (EtOH), which was further confirmed by crossover
experiments. Moreover, it was also confirmed that 10 and 12
did not interconvert under these reaction conditions (see
the Supporting Information for details). Interestingly, the di-
lution of the methanolic reaction increased the formation of
the monoamine 10ab up to 79% yield (entry 7). Further ex-
periments with the dibenzylamine derivatives 9b–d, which
share a similar steric environment at the nitrogen center but
feature different electronic properties,[13] showed a clear re-
lationship between the nucleophilicity of the amine and the
ratio of monoamine 10/diamine 11 in the reaction mixtures
(Table 1, entries 7–9). The reaction with amine 9d, the worst
nucleophile, was very slow and it afforded monoamine 10ad
in 47% yield (50% of conversion) after 7 days at room tem-
perature. We did not observe other chemical entities differ-
ent to the products and starting materials in the crude reac-
tion mixture. This fact seems to point to an early rate limit-
ing step in this domino reaction.
Table 1. Domino reaction of diyne 1a and secondary amines 9a–d.[a]
Entry
Solvent
Amine
10
11
[%]
[%]
1
2
3
4
5
6
7
8
9
CH2Cl2
CH2Cl2
MeOH
EtOH
C4H8NH[b]
C4H8NH
C4H8NH
C4H8NH
Bn2NH
9a
9a
9a
9a
9b
9b
9b
9c
9d
30
9
67
>99
>99
86
53
21
33
7
n.d.[c]
n.d.
n.d.
10[e]
39
EtOH
MeOH
MeOH[d]
MeOH[d]
MeOH[d]
Bn2NH
Bn2NH
79
The chemical efficiency and the complexity generation
power of this reaction exceeded all our expectations. The re-
PMDBA[f]
PNDBA[g]
67
93[h]
À
À
action generated up to four new bonds (2C N, 1 C C, 1C=
O) and one ring with excellent atom economy and an opera-
tionally simple protocol.
[a] Reaction conditions: diyne (0.10 mmol), amine (0.22 mmol), solvent
(2 mL). [b] 0.11 mmol. [c] n.d. =none detected. [d] 0.01m. [e] Roughly
equimolecular mixture of monoamides 11ab and 12ab. [f] PMDBA=
The set of experimental data can be rationalized through
the O-enolate-driven domino manifold outlined in
bis(4-methoxyphenyl)methanamine.
[g] PNDBA=bis(4-nitrophenyl)-
ACHTUNGTRENNUNG
9572
ꢃ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 9571 – 9575