Organic Letters
Letter
After our subsequent experience in developing a successful
MHAT intermolecular coupling of Cbz hydrazones to access
amines10 as well as tosylhydrazones11 as a general alkylation
reaction, we decided to revisit the intermolecular MHAT
coupling reaction of aldehydes. Based on both experimental
observations and mechanistic considerations,12 we proposed a
new strategy involving the addition of FeII, which could play
multiple positive roles within the catalytic cycle of the reaction
(Figure 2). It was envisaged that the addition of FeII would
Table 1. Screening of Reaction Conditions
entry
1a
2a
FeIII/FeII
solvent
time
yield
1
2
3
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
0.2:0
1:0
1:0
EtOH
EtOH
EtOH
EtOH
24 h
24 h
24 h
24 h
24 h
24 h
24 h
24 h
24 h
24 h
8 h
7%
25%
10%
32%
48%
50%
70%
35%
69%
58%
69%
59%
a
4
5
6
7
0.2:0.8
0.2:0.8
0.5:0.5
0.5:0.5
0:1
0.5:0.5
0.5:0.5
0.5:0.5
0.2:0.2
b
THF
THF
THF
THF
THF
THF
THF
b
c
d
c
8
c
9
c
10
11
12
c
c
THF
8 h
a
b
c
Heated at 60 °C. 2 equiv of MeOH as additive. 10 equiv of MeOH
d
as additive. Open to air.
1a (entry 9). Unexpectedly, however, adding more alkene 2a
proved detrimental, likely because the alkene is the least stable
component in the reaction mixture and more prone to side
reactions before the coupling takes place (entry 10). Cutting
the reaction time from 24 to 8 h produced almost identical
results (entry 11), whereas a further reduction to 3 h proved
unfeasible. Reducing the quantity of Fe to substoichiometric
amounts (0.2:0.2 of FeIII/FeII) resulted in a respectable 59%
yield (entry 12). Extending the reaction time to compensate
for the expected loss of reactivity from using less catalyst led to
a lower rather than higher yield. Once the optimum reaction
conditions were established (Table 1, entries 7 and 11), we
began to explore the scope of the reaction (Scheme 1).
Modifying the alkene component revealed that the reaction
worked with a wide range of functional groups (2a−k)
(Scheme 1), although slight modifications of the reaction time
and equivalents were required for more optimal results (see
Scheme 1 footnotes). For example, the very low yield of 3g
under the optimized conditions shown in Table 1 was greatly
improved by increasing the reaction time to 48 h. On the other
hand, the presence of a Lewis basic substituent on the alkenes,
as in the synthesis of 3c and 3e,16 generally accelerated the
reaction, which was usually completed within 8 h. As might be
expected, more substituted alkenes fared worse, the increased
stability of the carbon-centered radical favoring the reverse
process.
Compound 3k needed extensive reaction optimization to
achieve a relatively moderate 40% yield, while 3l (derived from
the tertiary radical intermediate) gave a disappointingly low
13% yield.
Variations in the acceptor showed that aromatic aldehydes
(3m−s) with both electron-donating and -withdrawing
substituents are well tolerated. The results were strikingly
improved when the benzaldehyde counterpart incorporated an
oxygenated ortho substituent (OH or OMe) that can act as a
Lewis base in the reactions leading to compounds 3r and 3s.
Aromatic heterocycles were also feasible, such as thiophenes
(3t) or pyridines (3u), as were aliphatic aldehydes (3v).
Given the important role of FeII in the intermolecular
coupling reaction, we next sought to evaluate its effects on the
Figure 2. Possible beneficial roles of FeII in the MHAT coupling
reaction of nonactivated alkenes with aldehydes.
facilitate the SET process, enabling a faster reduction and rapid
entrapment of the formed alkoxyl radical, and thus pre-empt a
reverse reaction via β-fragmentation (option A). Second, by
reacting with the initially formed carbon radical species
(option B), FeII would stabilize the radical via the persistent
radical effect (P.R.E)12b,13 and prevent its loss before the
desired reaction could take place. Finally, FeII may act as a
Lewis acid, lowering the activation energy of the reaction8 and
subsequently facilitating a direct SET process (option C). We
report here the validation of this theory and the first successful
use of aldehydes as radical acceptors in MHAT reactions.
To evaluate the reaction parameters, based on our
intermolecular MHAT couplings of alkenes with hydrazones,10
4-cyanobenzaldehyde 1a was chosen as the acceptor group and
4-phenyl-1-butene 2a as the radical precursor. As can be seen,
the use of FeIII alone in EtOH gave the coupled product 3a in
very low yields (7%) (Table 1, entry 1), which were
moderately improved using stoichiometric quantities of
Fe(acac)3 in EtOH (entry 2). As in all our previously
developed intermolecular MHAT coupling reactions,10,11
heating was found to be detrimental (entry 3). We then
began to evaluate the effect of adding FeII to the reaction,
observing a slight increase in yield to 32% when using
stoichiometric iron in a 2:8 FeIII/FeII ratio (entry 4), although
the improvement was far less than expected.
However, after changing the solvent from EtOH to THF
with MeOH as an additive (2 equiv),14 a synthetically useful
yield (48%) was obtained for the first time (entry 5). Changing
the Fe ratio to 1:1 led to a minor improvement (entry 6). In
further tests, increasing the amount of MeOH to 10 equiv
resulted in a higher yield (entry 7), but 20 equiv led to only a
slight improvement compared to 2 equiv.
The use of FeII alone, open to the air, gave a lower yield
(entry 8), probably due to competing Mukaiyama oxidation.15
Next, when evaluating the acceptor 1a/donor 2a ratio, the
yield was found to be unaffected by increasing the amounts of
B
Org. Lett. XXXX, XXX, XXX−XXX