Regiodivergent access to five- and six-membered benzo-fused lactams: Ru-catalyzed olefin hydrocarbamoylation

Article abstract of DOI:10.1021/ja411913e

We report herein a new strategy of the Ru-catalyzed intramolecular olefin hydrocarbamoylation for the regiodivergent synthesis of five- and six-membered benzo-fused lactams starting from N-(2-alkenylphenyl)formamides. Using a combined catalyst of Ru₃

Full text of DOI:10.1021/ja411913e

Article
pubs.acs.org/JACS
Regiodivergent Access to Five- and Six-Membered Benzo-Fused
Lactams: Ru-Catalyzed Olefin Hydrocarbamoylation
Bin Li,†,‡,§ Yoonsu Park,‡,† and Sukbok Chang*,†,‡
†Center for Catalytic Hydrocarbon Functionalizations, Institute of Basic Science (IBS), Daejeon 305-701, Korea
‡Department of Chemistry, Korea Advanced Institute of Science & Technology (KAIST), Daejeon 305-701, Korea
§State Key Laboratory of Element-Organic Chemistry, College of Chemistry, Nankai University, Tianjin 300071, People’s Republic of
China
S
* Supporting Information
ABSTRACT: We report herein a new strategy of the Ru-catalyzed
intramolecular olefin hydrocarbamoylation for the regiodivergent syn-
thesis of five- and six-membered benzo-fused lactams starting from N-(2-
alkenylphenyl)formamides. Using a combined catalyst of Ru3(CO)12/
Bu4NI in DMSO/toluene cosolvent (catalytic system A), a 5-exo-type
cyclization proceeds favorably to form indolin-2-ones as a major product
in good to excellent yield. When the reaction was conducted in the
absence of halide additives in DMA/PhCl (catalytic system B), 3,4-
dihydroquinolin-2-ones were obtained in major in moderate to high yield
via a 6-endo cyclization process. An excellent level of regioselectivity was
observed with a variety of substrates to deliver 5-exo- or 6-endo-cyclized
lactams. It was found that while the selective cyclization was controlled primarily by the choice of catalytic systems employed, it
was also greatly influenced by the structural nature of substrates. A halide-bridged trinuclear complex [Ru3(CO)10(μ2-I)]− is
postulated to be an active species in the catalytic system A. Two reaction pathways are proposed, in which the Ru-catalyzed
oxidative addition of formyl C−H or N−H bond initiates the subsequent cyclization processes.
tively by nickel and Lewis acid.8c Recently, Cramer revealed an
INTRODUCTION
■
asymmetric version of this reaction using a novel chrial
Transition metal-catalyzed hydroacylation reaction,1 a process
diaminophosphine oxide.8a In addition, Carreira et al.
involving an insertion of unsaturated compounds (such as
developed a Ru-catalyzed 5-endo-cyclization of allylic forma-
alkenes,2 alkynes,3 or ketones4) into activated C−H bond of
mides to yield substituted pyrrolidones under CO atmospher-
aldehydes, constitutes one of the most economical and efficient
e.7a In this approach, only five-membered ring products were
methods for the C−C bond formation.5 It is surprising to
obtained even with substrates bearing homoallylic and
realize that whereas hydroacylation has been well studied, the
bishomoallylic tethers.
related hydroesterification6 and, more notably, hydrocarbamoy-
lation have been less explored. In particular, the latter case is
mainly due to the difficulty of insertion of suitable metals into
the formamido C−H bonds when compared to the
corresponding aldehyde bonds. As a result, only a few metal-
Despite these impressive advances, we envisioned that a
stereodivergent intramolecular hydroamidation would be highly
interesting and attractive to obtain nitrogen-containing hetero-
cycles when readily available N-(2-alkenylphenyl)formamides
mediated systems including ruthenium7 and nickel8 have been
are used as a common starting material. Herein, we present our
new development of the Ru-catalyzed selective cyclization
scrutinized for the hydrocarbamoylation. However, in the Ru-
procedures without requiring external CO atmosphere
catalyst systems, the developed procedures often require CO
atmosphere because of a competitive, irreversible decarbon-
(Scheme 1).
ylation of activated intermediates.7a,b,e In this regard, we
In this approach, it is especially noteworthy that two feasible
cyclization modes (5-exo vs 6-endo) can be controlled starting
reported that a chelation strategy could be utilized to effectively
from the same starting materials simply by tuning the catalytic
suppress the undesired decarbonylation of ruthenium acyl
conditions, thus giving rise to both indolin-2-one (2) and 3,4-
hydride intermediates eventually to achieve efficient intermo-
lecular hydrocarbamoylation reaction.7c
On the other hand, an intramolecular hydrocarbamoylation
process allows an atom-economical access to synthetically
attractive amido-containing heterocycles. Up to now, only a few
dihydroquinolin-2-one (3) derivatives. To the best of our
knowledge, this is the first example of regiodivergent cyclization
method in the intramolecular olefin hydrocarbamoylation
reaction.9 In needs to be emphasized that both of the resulting
methods have been reported for this purpose. In 2009, Nakao
and Hiyama described an elegant example of cyclization of
homoallylic formamides to give γ-lactams catalyzed coopera-
Received: November 22, 2013
Published: December 27, 2013
© 2013 American Chemical Society
1125
dx.doi.org/10.1021/ja411913e | J. Am. Chem. Soc. 2014, 136, 1125−1131

Journal of the American Chemical Society
Article
a
Scheme 1. Regiodivergent Intramolecular
Table 1. Optimization of Reaction Conditions
Hydrocarbamoylation of N-(2-Vinylphenyl)formamides To
Form Indolin-2-ones or 3,4-Dihydroquinolin-2-ones
Ru
additive
yield (%)
b
entry (mol %)
(mol %)
T (°C)
solvent
DMF
(2a:3a)
1
2
3
4
5
5
5
5
5
5
Bu4NI (15)
Bu4NI (15)
Bu4NI (15)
Bu4NI (15)
Bu4NI (15)
120
120
120
120
120
89 (5.8:1)
94 (8.4:1)
79 (12.2:1)
78 (13.1:1)
91 (9.1:1)
DMSO
dioxane
toluene
DMSO/toluene
(3:1)
6
5
4
0
4
4
4
4
4
4
Bu4NI (15)
Bu4NI (12)
Bu4NI (12)
Bu4NI (12)
Bu4NI (12)
120
120
120
100
140
120
120
120
120
DMSO/toluene
(1:1)
95 (10.9:1)
96 (11.1:1)
NR
c
7
DMSO/toluene
(1:1)
five- and six-membered benzo-fused lactam skeletons are core
structures of numerous natural products, marketed drugs (or
their candidates), and synthetic compounds as exemplified in
Figure 1.10,11 As a consequence, a new efficient and selective
method for the construction of benzo-fused lactams is highly
desirable.
c
c
c
c
c
c
8
9
DMSO/toluene
(1:1)
DMSO/toluene
(1:1)
36 (8.0:1)
96 (11.0:1)
93 (6.4:1)
90 (9.3:1)
88 (10.8:1)
72 (1:2.0)
10
11
12
13
DMSO/toluene
(1:1)
Bu4NCl
(12)
DMSO/toluene
(1:1)
Bu4NBr
(12)
DMSO/toluene
(1:1)
NaI (12)
DMSO/toluene
(1:1)
14
none
DMSO/toluene
(1:1)
15
16
17
18
19
5
5
5
5
5
none
none
none
none
none
120
120
120
120
120
DMSO
DMA
xylene
PhCl
84 (1:1.2)
96 (1:1.6)
85 (1:4.5)
82 (1:5.1)
89 (1:5.1)
DMA/PhCl
(1:5)
a
b
Reactions in 0.2 mmol scale (0.5 M). Combined yield and ratio
1
(2a:3a) of crude reaction mixture determined by H NMR (internal
standard: 1,1,2,2-tetrachloroethane). For 6 h. NR = No reaction.
c
Figure 1. Pharmaceuticals and natural products containing indolinone
(top) and dihydroquinolinone (bottom) structure.
lower amounts of catalyst (entries 6,7). On the other hand, no
desired product was obtained in the absence of ruthenium
catalyst (entry 8). The reaction efficiency was also sensitive to
the reaction temperatures (entry 9), while the selectivity was
not altered at higher temperature (entry 10). Whereas
ammonium salts of chloride or bromide resulted in slightly
decreased regioselectivity (entries 11,12), NaI displayed almost
a similar level of additive effects when compared to Bu4NI
(entry 13).
During the course of our studies, we found that the
regioselectivity was changed to an opposite direction, now
favoring a six-membered lactam (3a) in the absence of halide
additives (entry 14). We were pleased to see that the selectivity
was further increased upon modulating the solvent systems: the
6-endo cyclization is more favored in less polar solvents (entries
15−18). In this line, chlorobenzene (PhCl) turned out to be
one of the most effective solvents delivering 6-endo products
predominantly. In addition, the use of a cosolvent system
(DMA/PhCl, 1:5) brought about increased product yields
while maintaining the regioselectivity still high when compared
to that in PhCl only (entry 19). Although this selectivity for the
6-endo-type cyclization is not as high as for the 5-exo-route, this
RESULTS AND DISCUSSION
■
We initiated our optimization experiments of an intramolecular
hydrocarbamoylation reaction using N-(2-vinylphenyl)-
formamide 1a (Table 1)12 on the basis of our previous studies
of the Ru-catalyzed intermolecular hydroesterification and
hydroamidation of olefins and alkynes.6e,g−i,7c We were pleased
to find that the reaction indeed did proceed leading to cyclized
products. Using Ru3(CO)12 catalyst (5 mol %) and Bu4NI
additive (15 mol %), the conversion was smooth at 120 °C to
furnish a mixture of five- and six-membered lactams (2a and 3a,
respectively) in a high combined yield (entry 1). Interestingly,
the ratio of two isomeric products was found to be dependent
on solvents employed. For instance, it was increased to 8.4:1
favoring indolin-2-one (2a) in DMSO (entry 2). The selectivity
was further improved up to 12−13:1 in relatively nonpolar
solvents (dioxane or toluene), but sacrificing their reactivity to
some extents (entries 3,4). A cosolvent system turned out to be
most satisfactory in terms of both product yield and
regioselectivity, and an equal mixture of DMSO and toluene
was found to be most optimal, thus allowing the use of slightly
1126
dx.doi.org/10.1021/ja411913e | J. Am. Chem. Soc. 2014, 136, 1125−1131

Journal of the American Chemical Society
Article
result is significant in that both benzo-fused lactam isomers are
readily accessible at our will simply by tuning the catalytic
systems starting from the same starting materials. It is also
noteworthy that no external CO atmosphere was required in
the current regiodivergent intramolecular olefin hydrocarba-
moylation procedures.
With the promising optimal conditions, we first examined the
scope of N-(2-vinylphenyl)formamides with diverse arene
substituents for the synthesis of indolinones under the catalytic
system A (Table 2). The cyclization proceeded smoothly
when compared to the electron-donating groups at the same
position. For instance, whereas substrates with 4-alkoxy and
phenoxy substituents were cyclized to produce indolin-2-ones
with the ratio of 8.4−9.8:1 (entries 6−8), it became more
selective (11.4−13.5:1) from starting materials having halide
groups at the same position (entries 9−11). Various functional
groups commonly encountered in organic synthesis were well
tolerated, such as alkoxy (entries 4−5), phenoxy (entry 6),
halides (entries 7−9 and 16), ester (entry 12), cyano (entry
13), trifluoromethyl (entry 14), and ketone (entry15). A
substrate bearing two substituents was readily cyclized with
high efficiency (entry 16). In addition, a reaction of a
naphthalene substrate (1q) was facile under the catalytic
system A leading to the corresponding product (2q) in
excellent yield and selectivity (entry 17).
a
Table 2. Substrate Scope for Indolin-2-one Synthesis
As the next step, we also investigated the substrate scope for
the synthesis of 3,4-dihydroquinolin-2-ones under the catalytic
system B that do not employ halide additives (Table 3). In
general, this 6-endo-type cyclization took place in good yields
and modest to high selectivity. It was interesting to observe that
the selectivity (5-exo vs 6-endo) was changed dramatically
depending on the substitution position. For instance, while the
reaction occurred highly selectively (1:6.2−13.1) with sub-
Table 3. Substrate Scope in the Synthesis of 3,4-
a
Dihydroquinolin-2-ones
a
b
Reactions in 0.3 mmol scale. Combined yields of isolated products:
ratio of crude reaction mixture determined by 1H NMR (internal
standard: 1,1,2,2-tetrachloroethane).
irrespective of the electronic nature of substituents to afford the
desired products in good to excellent yields (68−96%) as well
as a high level of regioselectivity (8.4−13.5:1). Importantly,
from the synthetic point of view, the resulting two isomers
could be easily separated by silica gel column chromatography.
The position of arene substituents little affected the reaction
efficiency and selectivity as demonstrated representatively by a
methyl group (entries 2−5). On the other hand, the selectivity
for the formation of five-membered lactams (2) was slightly
higher with substrates having electron-deficient substituents
a
b
Reactions in 0.3 mmol scale. Combined yield of isolated products:
ratio of crude reaction mixture determined by 1H NMR (internal
standard: 1,1,2,2-tetrachloroethane).
1127
dx.doi.org/10.1021/ja411913e | J. Am. Chem. Soc. 2014, 136, 1125−1131

Journal of the American Chemical Society
Article
a
strates bearing a methyl group at the 3-, 4-, and 5-positions
(entries 2−4), the cyclization was not selective when it is
located at the 6-position (entry 5). This result is in a stark
contrast to the 5-exo-cyclization process wherein the selectivity
was maintained high regardless of the substitution position
(Table 2, entries 2−5).
Table 4. Scope of Substrates Bearing Various Olefins
The notable difference of selectivity between two cyclization
modes is noteworthy providing a hint for the mechanistic
description (vide infra). In addition, the favorable formation of
six-membered lactams under catalytic system B was more
pronounced with substrates bearing electron-donating sub-
stituents. For example, a high level of selectivity for the 6-endo-
cyclization process was observed when alkoxy or phenoxy
groups were substituted (entries 6−8), whereas it was
decreased with halide-containing substrates (entries 9−11).
Functional group tolerance was found to be excellent under the
present catalytic system B as in the case of the 5-exo-cyclization.
Next, we scrutinized the scope of substrate variants
containing different olefins other than a vinyl group (Table
4). We were pleased to observe that the regiodivergent
synthesis was also possible with these derivatives to furnish
either 5-exo- or 6-endo-cyclized lactams with satisfactory
selectivity. The intramolecular hydrocarbamoylation proceeded
smoothly for the insertion into an E-propenyl group leading to
3-ethylindolidin-2-one (2r) and 3-methyl-3,4-dihydroquinolin-
2-one (3r) as major products under catalytic systems A and B,
respectively (entries 1−2). A similar trend with regard to
reactivity and selectivity was observed with a substrate (1s)
bearing 4-chloro substituent (entries 3,4). In addition, when Z-
propenyl isomer (cis-1r) was allowed to react, the olefinic
geometry was found to little affect the outcome of reactivity
and selectivity (entries 5,6). It was interesting to compare the
pattern of selectivity in the cyclization of substrates bearing
vinyl and propenyl moieties, although the exact reason is not
clear at the current stage. While higher selectivity was observed
in the formation of indolin-2-ones (5-exo) from vinyl-
containing substrates when compared to 6-endo-cyclization
process (Table 2), cyclization of substrates bearing a propenyl
group proceeded with lower selectivity for the 5-exo-process
but with higher preference for the 6-endo-products (Table 3).
The introduction of a benzylic substituent at the vinyl
terminus (1t) displayed no detrimental effect on the reactivity
and selectivity in the 5-exo-process (entry 7). However, modest
efficiency and selectivity were observed when 1t was subjected
to the 6-endo-procedure (catalytic system B, entry 8). It was
found that a replacement of the vinyl terminus from benzyl (1t)
to phenyl group (1u) resulted in more distinctive outcomes.
Whereas the 5-exo-process took place exclusively under the
employed system A albeit in lower yield (entry 9), 1u was not
reactive under the 6-endo-cyclization procedure (entry 10).
The above observed selectivity pattern led us to postulate
that the hydrocarbamoylation process is sensitive more to the
steric environment of substrates particularly around the olefinic
moiety, thus giving rise to the regiodivergent routes. To test
this hypothesis, two additional substrates (1v and 1w) bearing
gem-disubstituted olefinic groups were prepared and subjected
to the cyclization conditions. As expected, whereas the 5-exo-
process was not selective (entry 11), 6-endo-cyclization of 1v
was much more favored to afford 4-methyl-3,4-dihydroquino-
lin-2-one (3v) with excellent selectivity (1:14.5, entry 12). This
behavior was shown more distinctly with a bulkier gem-alkenyl
substituent (1w). When 1w bearing α-phenylvinyl moiety was
subjected to the catalytic system B, only a six-membered lactam
a
Catalytic system A: 1 (0.3 mmol), Ru3(CO)12 (5 mol %), Bu4NI (15
mol %), DMSO/toluene (1:1, 0.5 M), 120 °C, 20 h. Catalytic system
B: 1 (0.3 mmol), Ru3(CO)12 (5 mol %), DMA/PhCl (1:4, 0.33 M),
120 °C, 24−36 h. Combined yield of isolated products: yield in
parentheses and ratio of crude reaction mixture determined by H
NMR. NMP was used as solvent, 36 h. DMA/PhCl (1:5, 0.33 M, 16
h), and 21% of trans-1r was also detected.
b
1
c
d
3w was exclusively obtained (entry 14), while selectivity of its
5-exo-process under catalytic system A was modest (entry 13).
To the best of our knowledge, this route accessible to 2v and
2w represents the first example of generating a quaternary
carbon center through an olefin hydrocarbamoylation reaction,
although the selectivity still remained moderate at the present
stage.
It was notable to observe that a substrate 1x having an allylic
double bond underwent the cyclization to afford 2r and 3r as
products (entries 15,16), which are the same products obtained
from the reaction of propenyl-substituted substrates (cis- and
trans-1r, entries 1,2 and 5,6). Interestingly, seven-membered
compounds (4,5-dihydrobenzo-[b]-azepin-2-one) were not
1128
dx.doi.org/10.1021/ja411913e | J. Am. Chem. Soc. 2014, 136, 1125−1131

Journal of the American Chemical Society
Article
detected in this cyclization reaction. In addition, isomerization
of an allylic double bond to trans-propenyl olefin (from 1x to
trans-1r) was observed to occur in a notable extent (21% after
16 h) during the course of the cyclization (entry 16), which
provides a mechanistic implication that a metal hydride forms
under the employed catalytic conditions (vide infra).
To probe the working modes in the current cyclization
reaction, we conducted a series of experiments with isotopically
labeled substrates (Scheme 3).17 Under both catalytic
Scheme 3. Isotopic Studies with [D]C-1a
In our previous report,6e it was shown that the addition of
certain halides resulted in a dramatic enhancement in reaction
efficiency in the Ru-catalyzed intermolecular hydroesterification
of alkenes and alkynes.13 For example, ammonium iodide was
found to display significant improvement, whereas the effects
were less pronounced with the corresponding chloride or
bromide salts. In the present intramolecular hydrocarbamoyla-
tion, however, the type of halides does not seem to display
much distinctive influence on the reactivity, suggesting that the
catalytically active species might have similar structures
irrespective of halide additives.
It is known that a series of ruthenium complexes is generated
when a solution of Ru3(CO)12 is treated with halide ions
(Scheme 2).14 Notably, it is shown that the halide-bridged
conditions (A and B), the reaction of substrate [D]C-1a
bearing C−D at the formyl carbon afforded a five-membered
lactam with a high level of deuterium incorporation at the C3
and methyl moiety, while a six-membered isomer was
incorporated with a much lower deuterium extent (eqs I and
II). These results suggest that: (i) an initial insertion of the
reactive catalyst into a formyl C−H bond is involved under
both catalytic systems; and (ii) the subsequent olefin
hydrometalation process is reversible and nonselective. When
the reaction mixture was analyzed at a partial conversion, no
loss in the extent of deuterium contents at the formyl C−D
bond was observed (eq III). In addition, the olefinic double
bond of recovered [D]C-1a was not scrambled with any
deuterium. This result implies that the Ru-mediated cleavage of
a formyl C−D bond is presumably irreversible, leading to the
corresponding ruthenium acyl hydride species.
Scheme 2. Halide-Mediated Interconversion between
Ru3(CO)12 and Its Halide Complexes14a
The direct activation of a formamido C−H bond in the
present hydrocarbamoylation reaction was also supported by
the cyclization of a tertiary amide substrate 1y, which afforded
the corresponding five-membered ring product 2y albeit in low
yield (Scheme 4). This stands in contrast with the observation
complexes are readily interconverted and that the exact position
of equilibrium depends on both CO pressure and halide types
employed. For instance, Ru3(CO)12 reacts with chloride or
bromide ion to form initially [Ru3(CO)11(X)]− (4), which is
converted to a bridging species [Ru3(CO)10(μ2-X)]− (5a or
5b) and then to tetranuclear clusters [Ru4(CO)13(μ2-X)]− (6)
having a bridged halide ligand. In the case of iodide, direct
formation of a bridged complex 5c was observed, which is then
transformed to [Ru3(CO)9(μ3-I)]− (7) with a triply bridged
iodide ligand. Because the equilibrium between Ru3(CO)12 and
its monomeric Ru(CO)5 species needs high CO pressure,15 it is
reasonable to rule out a possibility that the added halide ions
dissociate the trinuclear precursor into a monomeric species
under the present reaction conditions. On the basis of these
results, we postulate that complexes 5 might be a common
catalytically active species responsible for the 5-exo-cyclization
process under catalytic system A.16
It needs to be empathized that the regioselectivity toward the
5-exo and 6-endo-cyclization is significantly altered by the
presence of halides, suggesting that a different catalytic cycle
might be operative by the additives in the present
carbamoylation. This stands in contrast to our previous Ru-
catalyzed hydroesterificaton,6e in which halide additives did not
change the regioselectivity for the formation of linear versus
branched isomeric esters.
Scheme 4. Reaction of a Tertiary Amide Substrate 1y
made by Carreira et al. in their Ru-catalyzed cyclization of
allylic formamides, in which the free amide N−H bond was
essential for the reaction progress.7a
The isotopic study was repeated, this time with a substrate
[D]N-1a containing a N−D bond (Scheme 5).17 Under both
catalytic systems A and B, deuterium was incorporated to a
significant extent in each five- and six-membered cyclic product
when the reaction was completed (eqs IV and V). This
observation suggests that Ru-mediated cleavage of the N−D
bond is also involved in the catalytic cycle and that the
subsequent olefin insertion step is reversible and nonselective.
When the reaction was interrupted at 70% conversion, the
1129
dx.doi.org/10.1021/ja411913e | J. Am. Chem. Soc. 2014, 136, 1125−1131

Journal of the American Chemical Society
Article
clearly favor one over the other at the present stage. As
described above, cleavage of formyl C−H bond by the active
ruthenium complex is believed to proceed irreversibly leading
to an acylruthenium hydride species A. Subsequent reversible
insertion of the olefin into the Ru−H bond can afford either
six-membered (B) or seven-membered ruthenacycle (C), which
undergoes a reductive elimination step to provide indolin-2-one
(2a) or 3,4-dihydroquinolin-2-one (3a), respectively. In
addition, a path leading to six-membered lactams is postulated
to be less favored as evidenced by the low deuterium
incorporation in the six-membered lactam from a reaction of
[D]C-1a (Scheme 3, eqs I and II).
Scheme 5. Isotopic Studies with [D]N-1a
On the other hand, an alternative pathway can also be
envisioned to operate, in which ruthenium cleaves an amido
N−H bond first leading to an amdioruthenium hydride species
D (Scheme 7). Subsequent olefin insertion into D may occur in
two directions leading to a five-membered ruthenacycle E or a
six-membered analogue F. At this stage, these ruthenacycles are
assumed to undergo β-hydride abstraction from a formyl C−H
bond to yield the corresponding isocyanate intermediates G or
H, respectively. In fact, Carreira et al. suggested that this route
would be responsible for their hydrocarbamoylation reaction of
allylic formamdies under similar ruthenium catalysis con-
ditions.7a Subsequent insertion of alkylruthenium moiety into
the in situ generated isocyanate group will deliver cyclic
imidates I or J, and then finally reductive cleavage and
tautomerization will release the desired products 2a or 3a with
the concomitant regeneration of catalyst. Again, it is proposed
that the 6-endo-cyclization process is favored under the catalytic
system B, although detailed studies are required.
recovered starting material (1a) was analyzed to have a
significant level of deuterium incorporation at the olefinic
double bond (eq VI), implying that the Ru-mediated cleavage
of a amido N−D bond is presumably reversible leading to the
corresponding ruthenium deuteride species. A large difference
in the deuterium incorporation in six-membered products from
two isotopic starting materials [D]C-1a and [D]N-1a can be
rationalized by taking an assumption that the 6-endo-cyclization
process proceeds more effectively via a N−H bond cleavage
pathway.
On the basis of the above observations and previous reports
by us and others,7 two reaction pathways are proposed for the
Ru-catalyzed olefin hydrocarbamoylation (Schemes 6 and 7).
Scheme 6. Proposed Reaction Pathway of Direct Activation
of the Formyl C−H Bond
Although more comprehensive mechanistic investigations are
required to present an unambiguous explanation for the
dichotomy between 5-exo and 6-endo pathways, we assume at
the present stage that halide additives play an important role by
changing the composition of catalytically active metal species.
In addition, we noticed that the choice of solvent also brought
about the change in selectivity in this regiodivergent cyclization.
CONCLUSIONS
■
In summary, we have developed the Ru-catalyzed intra-
molecular olefin hydrocarbamoylation of N-(2-alkenylphenyl)-
formamides for the regiodivergent synthesis of indolin-2-ones
and 3,4-dihydroquinolin-2-ones. Two convenient catalytic
conditions (A and B) were optimized for each type of isomeric
products. The reaction is featured to be atom-economical with
high efficiency and selectivity over a wide range of substrates.
Isotopic studies led us to propose two mechanistic pathways,
which differed in the way of ruthenium-mediated initial
cleavage of formyl C−H or amido N−H bond. While the
regioselectivity of the cyclization process is believed to be
primarily controlled by the nature of catalytic species generated
in situ, it is also dependent on the structural nature of the
substrates. Further studies on the synthetic applications and
mechanistic details are currently underway.
Scheme 7. Proposed Reaction Pathway of Initial Activation
of the N−H Bond
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedure and characterization of new com-
pounds (1H and 13C NMR spectra). This material is available
free of charge via the Internet at http://pubs.acs.org.
They differed in the way of ruthenium-mediated cleavage of
formyl C−H or amido N−H bond, although it is too early to
1130
dx.doi.org/10.1021/ja411913e | J. Am. Chem. Soc. 2014, 136, 1125−1131

Journal of the American Chemical Society
Article
41, 298. (c) Nakao, Y.; Idei, H.; Kanyiva, K. S.; Hiyama, T. J. Am.
Chem. Soc. 2009, 131, 5070.
AUTHOR INFORMATION
■
Corresponding Author
sbchang@kaist.ac.kr
(9) A recent review for the catalytic regiodivergent synthesis, see:
(a) Mahatthananchai, J.; Dumas, A. M.; Bode, J. W. Angew. Chem., Int.
Ed. 2012, 51, 10954. For recent examples, see: (b) Yang, Y.;
Buchwald, S. L. J. Am. Chem. Soc. 2013, 135, 10642. (c) Ding, Z.;
Yoshikai, N. Angew. Chem., Int. Ed. 2013, 52, 8574. (d) Shareef, A.-R.;
Sherman, D. H.; Montgomery, J. Chem. Sci. 2012, 3, 892. (e) Arndt,
Notes
The authors declare no competing financial interest.
̆
M.; Dindaroglu, M.; Schmalz, H.-G.; Hilt, G. Org. Lett. 2011, 13, 6236.
ACKNOWLEDGMENTS
■
For related examples of the hydroacylation reaction, see: (f) Coulter,
This research was supported by the Institute of Basic Science
(IBS) in Korea. B.L. is also grateful to the National Natural
Science Foundation of China (21372121) for financial support.
M. M.; Dornan, P. K.; Dong, V. M. J. Am. Chem. Soc. 2009, 131, 6932.
(g) Gonzalez-Rodríguez, C.; Pawley, R. J.; Chaplin, A. B.; Thompson,
́
A. L.; Weller, A. S.; Willis, M. C. Angew. Chem., Int. Ed. 2011, 50, 5134.
(10) For pharmaceuticals and natural products containing an
indolinone skeleton, see: (a) Galliford, C. V.; Scheidt, K. A. Angew.
Chem., Int. Ed. 2007, 46, 8748. (b) Marti, C.; Carreira, E. M. Eur. J.
Org. Chem. 2003, 2209. (c) Zhou, N.; Polozov, A. M.; O’Connell, M.;
Burgeson, J.; Yu, P.; Zeller, W.; Zhang, J.; Onua, E.; Ramirez, J.;
Palsdottir, G. A.; Halldorsdottir, G. V.; Andresson, T.; Kiselyov, A. S.;
Gurney, M.; Singh, J. Bioorg. Med. Chem. Lett. 2010, 20, 2658.
(d) Lozinskaya, N. A.; Sosonyuk, S. E.; Volkova, M. S.; Seliverstov, M.
Yu.; Proskurnina, M. V.; Bachurin, S. E.; Zefirov, N. S. Synthesis 2011,
273. (e) Chen, X.-H.; Wei, Q.; Luo, S.-W.; Xiao, H.; Gong, L.-Z. J. Am.
Chem. Soc. 2009, 131, 13819. (f) Grigg, R.; Whitney, S.; Sridharan, V.;
Keep, A.; Derrick, A. Tetrahedron 2009, 65, 4375. (g) Huang, A.;
Kodanko, J. J.; Overman, L. E. J. Am. Chem. Soc. 2004, 126, 14043.
REFERENCES
■
(1) For selected recent reviews, see: (a) Willis, M. C. Chem. Rev.
2010, 110, 725. (b) Park, Y. J.; Park, J.-W.; Jun, C.-H. Acc. Chem. Res.
2008, 41, 222. (c) Jun, C.-H.; Jo, E.-A.; Park, J.-W. Eur. J. Org. Chem.
2007, 1869. (d) Leung, J. C.; Krische, M. J. Chem. Sci. 2012, 3, 2202.
(2) For recent examples, see: (a) von Delius, M.; Le, C. M.; Dong, V.
M. J. Am. Chem. Soc. 2012, 134, 15022. (b) Chaplin, A. B.; Hooper, J.
F.; Weller, A. S.; Willis, M. C. J. Am. Chem. Soc. 2012, 134, 4885.
(c) Murphy, S. K.; Coulter, M. M.; Dong, V. M. Chem. Sci. 2012, 3,
355. (d) Piel, I.; Steinmetz, M.; Hirano, K.; Frohlich, R.; Grimme, S.;
̈
Glorius, F. Angew. Chem., Int. Ed. 2011, 50, 4983. (e) Zhang, H.-J.;
Bolm, C. Org. Lett. 2011, 13, 3900. (f) Phan, D. H. T.; Kou, K. G. M.;
Dong, V. M. J. Am. Chem. Soc. 2010, 132, 16354. (g) Coulter, M. M.;
Kou, K. G. M.; Galligan, B.; Dong, V. M. J. Am. Chem. Soc. 2010, 132,
16330. (h) Pawley, R. J.; Moxham, G. L.; Dallanegra, R.; Chaplin, A.
B.; Brayshaw, S. K.; Weller, A. S.; Willis, M. C. Organometallics 2010,
29, 1717. (i) Hirano, K.; Biju, A. T.; Piel, I.; Glorius, F. J. Am. Chem.
Soc. 2009, 131, 14190. (j) Shibata, Y.; Tanaka, K. J. Am. Chem. Soc.
2009, 131, 12552.
(3) For recent examples, see: (a) Sim, Y.-K.; Lee, H.; Park, J.-W.;
Kim, D.-S.; Jun, C.-H. Chem. Commun. 2012, 48, 11787. (b) Lenden,
P.; Entwistle, D. A.; Willis, M. C. Angew. Chem., Int. Ed. 2011, 50,
10657. (c) Parsons, S. R.; Hooper, J. F.; Willis, M. C. Org. Lett. 2011,
́
(h) Jossang, A.; Jossang, P.; Hadi, H. A.; Sevenet, T.; Bodo, B. J. Org.
Chem. 1991, 56, 6527. (i) Labroo, R. B.; Cohen, L. A. J. Org. Chem.
1990, 55, 4901.
(11) For pharmaceuticals and natural products containing an
indolinone skeleton, see: (a) Jones, G. In Comprehensive Heterocyclic
Chemistry II; Katritzky, A. R., Rees, C. W., Scriven, E. F. V., Eds.;
Pergamon Press: Oxford, UK, 1996; p 167. (b) Chen, M.-H.;
Fitzgerald, P.; Singh, S. B.; O’Neill, E. A.; Schwartz, C. D.; Thompson,
C. M.; O’Keefe, S. J.; Zaller, D. M.; Doherty, J. B. Bioorg. Med. Chem.
Lett. 2008, 18, 2222. (c) Harmata, M.; Hong, X. Org. Lett. 2007, 9,
2701. (d) Turner, K. L.; Baker, T. M.; Islam, S.; Procter, D. J.;
Stefaniak, M. Org. Lett. 2006, 8, 329. (e) Uchida, R.; Imasoto, R.;
Shiomi, K.; Tomoda, H.; Omura, S. Org. Lett. 2005, 7, 5701.
(f) Oshiro, Y.; Sakurai, Y.; Sato, S.; Kurahashi, N.; Tanaka, T.; Kikuchi,
T.; Tottori, K.; Uwahodo, Y.; Miwa, T.; Nishi, T. J. Med. Chem. 2000,
43, 177. (g) Semba, J.; Watanabe, A.; Kito, S.; Toru, M.
Neuropharmacology 1995, 34, 785. (h) Morita, S.; Irie, Y.; Saitoh, Y.;
Kohri, H. Biochem. Pharmacol. 1976, 25, 1836.
́
13, 998. (d) Gonzalez-Rodríguez, C.; Parsons, S. R.; Thompson, A. L.;
Willis, M. C. Chem.Eur. J. 2010, 16, 10950 and also see ref 2a.
(4) (a) Khan, H. A.; Kou, K. G. M.; Dong, V. M. Chem. Sci. 2011, 2,
407. (b) Phan, D. H. T.; Kim, B.; Dong, V. M. J. Am. Chem. Soc. 2009,
131, 15608 and references cited therein.
(5) (a) Trost, B. M. Acc. Chem. Res. 2002, 35, 695. (b) Trost, B. M.
Science 1991, 254, 1471.
(12) See the Supporting Information for details of the optimization
study.
(6) For reviews, see: (a) Brennfuehrer, A.; Neumann, H.; Beller, M.
ChemCatChem 2009, 1, 28. (b) El Ali, B.; Alper, H. Hydro-
carboxylation and hydroesterification reactions catalyzed by transition
metal complexes. In Transition Metals for Organic Synthesis, 2nd ed.;
Beller, M., Bolm, C., Eds.; Wiley-VCH: Weinheim, 2004; pp 113−132.
For selected examples, see: (c) Konishi, H.; Ueda, T.; Muto, T.;
Manabe, K. Org. Lett. 2012, 14, 4722. (d) Katafuchi, Y.; Fujihara, T.;
Iwai, T.; Terao, J.; Tsuji, Y. Adv. Synth. Catal. 2011, 353, 475. (e) Park,
E. J.; Lee, J. M.; Han, H.; Chang, S. Org. Lett. 2006, 8, 4355. (f) Wang,
L.; Floreancig, P. E. Org. Lett. 2004, 6, 4207. (g) Ko, S.; Lee, C.; Choi,
M.-G.; Na, Y.; Chang, S. J. Org. Chem. 2003, 68, 1607. (h) Na, Y.; Ko,
S.; Hwang, L. K.; Chang, S. Tetrahedron Lett. 2003, 44, 4475. (i) Ko,
S.; Na, Y.; Chang, S. J. Am. Chem. Soc. 2002, 124, 750. (j) Lugan, N.;
(13) For selected reviews on the halide effects in transition metal
catalysis, see: (a) Fagnou, K.; Lautens, M. Angew. Chem., Int. Ed. 2002,
41, 26. (b) Lavigne, G. Eur. J. Inorg. Chem. 1999, 917. For selected
examples revealing that the addition of certain salts to a solution of
rutheniumcarbonyl complexes results in catalytically more active
species, see: (c) Dombek, B. D. J. Organomet. Chem. 1989, 372, 151.
(d) Dombek, B. D. J. Organomet. Chem. 1983, 250, 467. (e) Knifton, J.
F. J. Am. Chem. Soc. 1981, 103, 3959. (f) Dombek, B. D. J. Am. Chem.
Soc. 1981, 103, 6508.
(14) (a) Rivomanana, S.; Lavigne, G.; Lugan, N.; Bonnet, J.-J.
Organometalllics 1991, 10, 2285. (b) Han, S.-H.; Geoffroy, G. L.;
Dombek, B. D.; Rheingold, A. L. Inorg. Chem. 1988, 27, 4355.
(15) (a) Koelliker, R.; Bor, G. J. Organomet. Chem. 1991, 417, 439.
(b) Bor, G. Pure Appl. Chem. 1986, 58, 543.
Lavigne, G.; Soulie,
Organometallics 1995, 14, 1712. (k) Lavigne, G.; Lugan, N.; Kalck, P.;
Soulie, J. M.; Lerouge, O.; Saillard, J. Y.; Halet, J. F. J. Am. Chem. Soc.
́
J. M.; Fabre, S.; Kalck, P.; Saillard, J. Y.; Halet, J. F.
́
(16) It can be envisioned that an asymmetric 5-exo cyclization would
be possible in principle because a stereogenic center is generated in 5-
exo products. However, as can be seen in the mechanistic proposal, the
ruthenium metal center under the catalytic system A does not form
any chiral environment even in the presence of chiral ammonium
halides (Scheme 2, complexes 5a−c).
1992, 114, 10669.
(7) (a) Armanino, N.; Carreira, E. M. J. Am. Chem. Soc. 2013, 135,
6814. (b) Nath, D. C. D.; Fellows, C. M.; Kobayashi, T.; Hayashi, T.
Aust. J. Chem. 2006, 59, 218. (c) Ko, S.; Han, H.; Chang, S. Org. Lett.
2003, 5, 2687. (d) Kondo, T.; Okada, T.; Mitsudo, T.-a. Organo-
metallics 1999, 18, 4123. (e) Tsuji, Y.; Yoshii, S.; Ohsumi, T.; Kondo,
T.; Watanabe, Y. J. Organomet. Chem. 1987, 331, 379.
(17) See the Supporting Information for details of experiments with
isotopically labeled substrates.
(8) (a) Donets, P. A.; Cramer, N. J. Am. Chem. Soc. 2013, 135, 11772.
(b) Miyazaki, Y.; Yamada, Y.; Nakao, Y.; Hiyama, T. Chem. Lett. 2012,
1131
dx.doi.org/10.1021/ja411913e | J. Am. Chem. Soc. 2014, 136, 1125−1131