DOI: 10.1039/C4CC08203F
ChemComm
COMMUNICATION
Journal Name
Table 1. Optimization of the reaction conditions.a
Entry
1b
Light source
dark
Catalyst [Ru]
--
1 mol%
--
1 mol%
1 mol%
1 mol%
Time (h)
Isolated yield (%)
32
30
18
5
5
4
16
21
23
94
92
97
2b
dark
3b
visible light
visible light
visible light
sunlight
4b
5c
6c
a All reactions were carried out using 2a (0.2 mmol), 3a (0.4 mmol), and
Pri2NEt (0.4 mmol) in deaerated methanol (10 mL) at rt. b The reaction was
carried out under argon atmosphere. c The reaction was carried out under air
atmosphere.
Scheme 3. The prospective mechanism
other dibromides 2b and 2c were applied in the
photocyclopropanation of 3a, which gave the desired products 4ba
and 4ca in excellent isolated yields, respectively (entries 20 and 21,
Table 2). Finally, electron-rich alkene 3t, aliphatic alkene 3u, and 2-
methylenemalononitrile 3v were applied under Condition A. No
cyclopropanes were formed (Scheme S1 in ESI). In all the above cases,
be highly nucleophilic for the following transformation. Thirdly, the
steric hindrance at α-position should be very little to enable the three-
membered ring formation process.
With these considerations in mind, diethyl 2,2-dibromomalonate 2a
and 2-benzylidenemalononitrile 3a were chosen as the model
substrates to examine this new cyclopropanation protocol. To our
delight, the reaction underwent smoothly under our initial attempt.
Then a series of control reactions were examined. Without light,
neither with nor without the catalyst, this reaction proceeded slowly
to give very low yields of the cyclopropane 4aa after full conversion of
no reduced 3 was formed.
Mechanistically, a set of reactions were performed to gain more
information. Firstly, radical scavenger TEMPO was added to this
reaction system. The addition of 1 equiv. of TEMPO resulted in a
slightly decreased yield. Further increase the amount of TEMPO to 5
equiv. led to similar results. In both cases, no obvious changes in the
reaction speed were observed. Next, 5 equiv. of BHT was used instead
of TEMPO, similarly, no inhibition was observed (Scheme S2 in ESI).
Considering that no obvious feature of free radical was observed,
since O2 in air, TEMPO, or BHT could not significantly affect this
transformation, radical species might not be involed in the rate-
determing step.15,16 This might mean that the reduction of carbon
radical into carbanion proceeds at an extremely rapid rate. Then a
dark reaction after irradiation for 5 minutes was also carried out
yielding only 29% of 4aa, which ruled out a radical chain mechanism,
since it gave a similar yield of 4aa with the reaction under Condition A
without light (Scheme S3 in ESI).
the starting material 2a (entries
1 and 2, Table 1). Under the
irradiation of visible light and in the absence of catalyst, a similar low
yield of 4aa was formed (entry 3, Table 1). When irradiated by visible
light and in the presence of 1 mol% of the photocatalyst, the reaction
speed dramatically increased, affording 4aa in 94% isolated yield
(entry 4, Table 1). Notably, this reaction could be carried out in air,
yielding 4aa in almost the same excellent yield (entry 5, Table 1). As
the time was shortened to approximately 5 hours, it enables this
transformation to be conducted under sunlight. Indeed, higher
efficiency and better yield were observed (entry 6, Table 1). So
Condition A (2 equiv.
3
, 2 equiv. Pri2NEt, 1 mol% Ru(bpy)3Cl2·6H2O,
methanol, visible light, and rt) and Condition B (2 equiv.
3
, 2 equiv.
Pri2NEt, 1 mol% Ru(bpy)3Cl2·6H2O, methanol, sunlight, and rt) were
applied for the following studies.14
Next, alkenes
8 with different configurations were employed to
investigate the stereochemistry of this cyclopropanation. As shown in
With the optimal conditions in hand, we investigated the scope of
Scheme 4, under the standard conditions, reactions of 2a with either
this photocyclopropanation with
a series of dibromomalonate
Z
-8 or
Notably, no cis-9 was formed in either case. To check whether there
was an isomerization of -8 into -8 under this reaction conditions,
the unconsumed reactant -8 and
-8 were recovered. Careful 1H
E-8 gave the same trans-product 9 in nearly the same yield.
derivatives and alkenes. Reactions under Conditions A and B gave
similar results (Table 2). Firstly, the electronic effect of substituents on
the phenyl ring of the alkenes was studied carefully. With strong
electron-withdrawing groups, such as methoxycarbonyl (entry 1,
Table 2), trifluoromethyl (entry 2, Table 2), and nitro (entries 3-5,
Table 2), the corresponding cyclopropane derivatives were formed in
good to excellent yields Excellent yields were also obtained for weak
electron-withdrawing groups substituted substrates, like fluorine
(entry 6, Table 2) and chlorine (entry 7, Table 2), and weak electron-
donating groups, like alkyl (entries 9-13, Table 2) and phenyl (entry 14,
Table 2). For substrates bearing strong electron-donating groups, like
alkoxy (entries 15-18, Table 2) and acetoxy (entry 19, Table 2), the
yields decreased slightly. The electronic effect of the substrate
strongly suggested a carbanion intermediated mechanism.11 Then
Z
E
Z
E
NMR analysis proved that there was no change in their double bond
configuration in the recovered starting material. These results suggest
a free bond rotation occured at the cyclization procedure, which fits
well with our designed reaction pathway via Michael addition and
subsequent intramolecular nucleophilic attraction.
Although the above results fitted well with our initially proposed
mechanism as shown in Scheme 3, some other possibilities of reaction
pathways should also be considered. (For further discussion, see ESI)
In summary, we have developed a visible light induced generation
of carbanion to achieve cyclopropanation of dibromomalonates with
2 | J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 2012