Imidazole derivatives as accelerators for ruthenium-catalyzed hydroesterification and hydrocarbamoylation of alkenes: Extensive ligand screening and mechanistic study

Article abstract of DOI:10.1002/cctc.201402986

Imidazole derivatives are effective ligands for promoting the [Ru₃(CO)₁₂]-catalyzed hydroesterification of alkenes using formates. Extensive ligand screening was performed to identify 2-hydroxymethylated imidazole as the optimal ligand. Neither carbon monoxide gas nor a directing group was required, and the reaction also showed a wide substrate generality. The Ru-imidazole catalyst system also promoted intramolecular hydrocarbamoylation to afford lactams. A Ru-imidazole complex was unambiguously analyzed by X-ray crystallography, and it had a trinuclear structure derived from one [Ru₃(CO)₁₂] and two ligands. This complex was also successfully used for hydroesterification. The mechanism was examined in detail by using D- and ¹³C-labeled formates, indicating that the hydroesterification reaction proceeds by a decarbonylation-recarbonylation pathway. Effective imidazole assistant: [Ru₃(CO)₁₂]-catalyzed hydroesterification of alkenes by using formates is drastically accelerated by imidazole derivatives and exhibits a broad substrate scope for both alkenes and formates. The Ru-imidazole complex also catalyzes the intramolecular hydrocarbamoylation of alkenes.

Full text of DOI:10.1002/cctc.201402986

DOI: 10.1002/cctc.201402986
Full Papers
Imidazole Derivatives as Accelerators for Ruthenium-
Catalyzed Hydroesterification and Hydrocarbamoylation of
Alkenes: Extensive Ligand Screening and Mechanistic
Study
Hideyuki Konishi,[a] Takashi Muto,[a] Tsuyoshi Ueda,[a, b] Yayoi Yamada,[a] Miyuki Yamaguchi,[a]
and Kei Manabe*[a]
Imidazole derivatives are effective ligands for promoting the
[Ru3(CO)12]-catalyzed hydroesterification of alkenes using for-
mates. Extensive ligand screening was performed to identify 2-
hydroxymethylated imidazole as the optimal ligand. Neither
carbon monoxide gas nor a directing group was required, and
the reaction also showed a wide substrate generality. The Ru–
imidazole catalyst system also promoted intramolecular hydro-
carbamoylation to afford lactams. A Ru–imidazole complex
was unambiguously analyzed by X-ray crystallography, and it
had a trinuclear structure derived from one [Ru3(CO)12] and
two ligands. This complex was also successfully used for hydro-
esterification. The mechanism was examined in detail by using
D- and 13C-labeled formates, indicating that the hydroesterifi-
cation reaction proceeds by a decarbonylation–recarbonylation
pathway.
Introduction
Esters are simple functional groups ubiquitously found in both
natural and synthesized compounds. As nature uses enzymes
to synthesize esters, chemists can use acid catalysts or conden-
sation reagents to obtain esters from the corresponding car-
boxylic acids and alcohols.[1] New synthetic methods for insert-
ing ester groups at the desired position in a target molecule
are needed. Alkenes, an important class of compounds, are
substrates for metal-catalyzed alkoxycarbonylation using
carbon monoxide (CO) and alcohols to synthesize esters.[2]
Since its discovery by Reppe,[3] this method has been tremen-
dously improved, enabling the synthesis of esters from alkenes
on an industrial scale. Later, metal-catalyzed hydroesterification
of alkenes using formates was developed.[4–8] This modification
of Reppe’s method allows the direct construction of esters
without using the toxic CO gas.[9] This method utilizes the pre-
existing ester group of formate instead of CO and an alcohol
to afford esters that are one-carbon-homologated products of
the original alkene. This method is especially useful if the prod-
uct is difficult to synthesize from a carboxylic acid and an alco-
hol because of, for example, acid lability or safety and prepara-
tion cost. Alkenes are readily available through various func-
tional group manipulations, and formate esters are inexpen-
sive, less toxic, and stable compounds. Several formate esters
are commercially available with an acceptable cost. Moreover,
catalytic hydroesterification is atom economical[10] and matches
the current trend to reduce waste production in synthetic pro-
tocols.
Pioneering work on metal-catalyzed hydroesterification
using formates was reported by Sneeden et al. in 1983.[4] How-
ever, this reaction was not practical because of the harsh con-
ditions employed and the use of methyl formate as a solvent.
Strict limitation of the substrate scope was also a major issue,
because only methyl formate and ethylene could be applied.
Although scattered examples[5] have appeared since the origi-
nal report, a general method for metal-catalyzed hydroesterifi-
cation was not published until 2002. Chang et al. demonstrat-
ed that formate substituted with a 2-pyridylmethyl group was
effective for the Ru-catalyzed hydroesterification of alkenes.[6]
This method was broadly applicable to alkenes; however, the
formate scope was limited to 2-pyridylmethyl formate. Recent-
ly, a Pd-catalyst was reported to be effective for hydroesterifi-
cation, but the formate scope was rather limited.[7] Although
hydroesterification with formates is an attractive method to
synthesize esters, development of a general method for both
alkenes and formates remains challenging.
[a] Dr. H. Konishi, T. Muto, Dr. T. Ueda, Y. Yamada, Dr. M. Yamaguchi,
Prof. Dr. K. Manabe
School of Pharmaceutical Sciences
University of Shizuoka
52-1 Yada, Suruga-ku, Shizuoka 422-8526 (Japan)
Fax: (+81)54-264-5586
E-mail: manabe@u-shizuoka-ken.ac.jp
[b] Dr. T. Ueda
Process Technology Research Laboratories
Pharmaceutical Technology Division
Daiichi Sankyo Co. Ltd.
1-12-1 Shinomiya
In the process of developing practical synthetic reactions for
biologically active compounds, we recognized the potential of
catalytic hydroesterification. Despite its lack of generality, sys-
Hiratsuka, Kanagawa 254-0014 (Japan)
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201402986.
ChemCatChem 0000, 00, 0 – 0
1
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
tematic screening of additives led us to the development of
highly effective imidazole derivatives for the reaction. Herein,
we report the development of a Ru-catalyzed hydroesterifica-
tion of alkenes using formates by adding imidazole derivatives
as accelerators.[8] The reaction has a wide substrate generality
in both alkenes and formates. Further application of these find-
ings led to the development of a Ru-catalyzed hydrocarbamoy-
lation of alkenes using formamides. Furthermore, mechanistic
studies for the catalytic hydroesterification reaction imply that
the reaction involves both decarbonylation and recarbonyla-
tion pathways.
Table 1. Catalytic
use
of
additives
for
hydroesterification.
Entry
Additive
Yield
[%][a]
3aa:3ab[b]
1
2
3
4
–
0
43
68
47
–
Results and Discussion
L1
L2
L3
73:27
35:65
75:25
Effect of ligands in Ru-catalyzed hydroesterification
Our investigations began by considering the intermediate of
Chang’s Ru-catalyzed hydroesterification using 2-pyridylmethyl
formate.[6] The stability of this intermediate, generated after ox-
idative addition of the formyl CÀH bond to Ru metal, presuma-
bly results from the coordination of the pendant pyridine ni-
trogen of the formate to Ru. At this point, we hypothesized
that an alcohol exchange of a simple formate with 2-pyridyl-
methanol (L1)[11] would form 2-pyridylmethyl formate
(Scheme 1), which was known as the active substrate for
[a] Combined isolated yields of 3aa and 3ab. [b] Determined by 1H NMR
spectroscopy.
thermore, imidazole L2 bearing a hydroxymethyl moiety was
more effective (entry 3).[12] Surprisingly, imidazole L3, which
lacked a hydroxy group, also afforded the product (entry 4).
These results indicate that an ester exchange is not necessary
for the reaction. These additives might function as ligands to
form an active Ru complex.
These initial studies led us to investigate the effect of li-
gands in greater detail. We screened various ligands shown in
Figure 1 for the hydroesterification. The ligands are catego-
rized into four classes: imidazoles with alkyl and aryl groups,
imidazoles with a hydroxy group, imidazoles with functional
groups, and other ligands.
The results of the screening are shown in Figure 2. Imida-
zoles with alkyl and aryl groups (L4–L18) generally afforded
the desired product in moderate yields, except for N-unsubsti-
tuted imidazole L4. For imidazoles with a hydroxy group (L2,
L19–L30), the yields largely depended on the ligand structures.
Among them, L21 gave the best combined yield (80% yield).
Uses of imidazoles with functional groups (L3, L31–L37) and
other ligands (L1, L38–L56) resulted in poor to modest yields
that did not exceed 60%. In the case of L53–L55, significant
amounts of benzyl alcohol were generated as a byproduct
(59–100% based on 1a).[13]
In most cases of the ligand screening, linear isomer 3aa was
preferentially obtained over branched isomer 3ab. However,
interesting reversal of regioselectivity was observed for N-sub-
stituted, hydroxymethylated imidazoles L2, L20, and L21, and
3ab was obtained as the major isomer. Increasing the steric
bulk at the hydroxymethyl carbon (L22–L26) or introduction of
substituents at C4 and C5 positions (L28, L29) resulted in
higher linear selectivity. Although the reason for the change in
regioselectivity is unclear, these results suggest that the regio-
selectivity of the catalytic hydroesterification reaction can be
controlled by the choice of ligand.
Scheme 1. Initial working hypothesis for Ru-catalyzed hydroesterification.
Chang’s hydroesterification. After the hydroesterification step,
a second alcohol exchange with the alcohol generated from
the first ester exchange would give the desired product and
regenerate L1. This reaction should require only a catalytic
amount of L1 and utilize, in principle, any formate ester if the
exchange occurs. Based on this working hypothesis, we investi-
gated the effects of catalytic amounts of additives such as L1
for the Ru-catalyzed hydroesterification reaction.
Then, we briefly tested the capability of L1 and similar com-
pounds as catalytic promoters using [Ru3(CO)12] as the catalyst
for the hydroesterification of 4-methoxystyrene (2a) with
benzyl formate (1a) in mesitylene (Table 1). Without additive,
the reaction did not proceed at all (Table 1, entry 1). However,
a catalytic amount of L1 indeed afforded the desired product
as a mixture of the linear and branched isomers (entry 2). Fur-
We optimized the reaction conditions using the best ligand
(L21), and finally found that the reaction using a slight excess
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
2
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Figure 1. Ligands screened for hydroesterification.
of alkene 2a under solvent-free
conditions afforded the desired
product
in
89%
yield
(Scheme 2). Interestingly, forma-
tion of the linear isomer was
slightly preferred.
Substrate scope for hydroester-
ification
An extensive study of the ligand
effects for the Ru-catalyzed hy-
droesterification of alkenes using
formates revealed that L21 was
optimal in terms of the desired
product yield. Therefore, L21
was used to investigate the sub-
strate scope of the reaction.
Figure 2. Screening of ligands for hydroesterification of 2a with 1a. y axis: combined isolated yield of 3aa and
3ab. Gray: yield of 3aa. Black: yield of 3ab.
First, the substrate scope of
formates
was
investigated
(Table 2). Benzyl formates with
substitution on the benzene ring afforded products in moder-
ate to high yield (Table 2, entries 1 and 2). 1-Naphthylmethyl
formate (1d) gave the product in high yields (entry 3). Aryl for-
mates afforded lower yields, and the formation of phenols was
observed (entries 4 and 5). An excess of ethyl formate (1g) and
a higher temperature were required to obtain the ethyl ester
Scheme 2. Hydroesterification using L21 under solvent-free conditions.
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
3
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
alkene, trans-stilbene (2i), re-
quired a higher temperature to
form the product, probably be-
cause of the steric hindrance. On
the other hand, b-methylstyrene
(2j) reacted at 1358C, but gave
three products (entry 9). This
was ascribed to alkene isomeri-
zation by the Ru catalyst[14] to
form allylbenzene (2k) followed
by hydroesterification of both 2j
and 2k. The same products were
obtained for the reaction of 2k
because it could also isomerize
to the thermodynamically more
stable 2j (entry 10). This result is
consistent with a report by Car-
reira et al., who recently demon-
strated the tandem Ru-catalyzed
isomerization and hydroesterifi-
cation of alkenes using allyl-
amines as substrates.[15] Overall,
both reactivity and regioselectiv-
ity in the Ru-catalyzed hydroes-
terification were strongly influ-
enced by the steric environment
of alkenes, with aromatic alkenes
being slightly more reactive than
the aliphatic ones. Nevertheless,
enhancement of the catalytic
ability of Ru by the imidazole
ligand was prominent.
Table 2. Substrate scope of formates.[a]
Entry
R
Product
Yield
[%][b]
83
(53:47)
1
2
61
(51:49)
79
(44:56)
3
Ph
1e
43
(68:32)
4
Ar[c]
1 f
57
(48:52)
5
Et
1g
48
(51:49)
6[d]
nC7H15
1h
76
(61:39)
7
Ph2CH
1i
73
(83:17)
8
9
H
0
Hydroesterification catalyzed
by a Ru–imidazole catalyst was
further applied to intramolecular
reactions (Table 4). Catalytic in-
tramolecular hydroesterification
may be one of the most efficient
methods to obtain lactones, al-
though this process had not
[a] Reactions were conducted with 1 (0.40 mmol), 2a (1.5 equiv.), [Ru3(CO)12] (5 mol%), and L21 (15 mol%)
under solvent-free conditions at 1358C for 24 h. [b] Combined isolated yields of the linear and branched iso-
mers. The ratio of isomers (determined by 1H NMR spectroscopy) is shown in parentheses. [c] Ar=4-methoxy-
phenyl. [d] The reaction was conducted on a 1.0 mmol scale at 1558C. 3.0 equiv. of 1g were used.
product (entry 6). Also, formates with a long alkyl chain 1h or
a bulky benzhydryl group 1i were applicable in this reaction
(entries 7 and 8). These results clearly indicate that hydroesteri-
fications using a Ru–imidazole catalyst have a wide substrate
generality for formates. The use of formic acid instead of for-
mates did not give the desired carboxylic acid at all (entry 9).
Next, the alkene scope was examined (Table 3). Monosubsti-
tuted terminal alkenes were good substrates for the reaction
(Table 3, entries 1 and 2). Geminally disubstituted alkene 2d af-
forded a single regioisomer in moderate yield (entry 3). The re-
activity and regioselectivity were affected by steric hindrance,
as observed in the reaction of limonene (2e), which afforded
the least sterically hindered isomer (entry 4). The reactivity of
cyclic alkenes depended on their structure. Although cyclooc-
tene (2 f) afforded the desired product in moderate conversion,
both norbornene (2g) and indene (2h) gave the correspond-
ing products in high yields (entries 5–7). An acyclic internal
been demonstrated until our initial report.[8a,16] Several sub-
strates bearing both formyl and alkenyl groups were readily
synthesized and applied to this reaction. As with intermolecu-
lar hydroesterification, imidazole ligands were necessary to
promote catalytic intramolecular hydroesterification (Table 4,
entry 1). Ligand L9, which lacks the hydroxymethyl group, pro-
duced better results than L21, which enabled the formation of
six- and seven-membered lactones from benzyl formate deriva-
tives (entries 2 and 3). Phenyl formate derivatives were suc-
cessfully applied by using L9. No electronic effects were ob-
served for different substituents on the benzene ring (en-
tries 4–6). Substrates 4e and 4 f gave the same products. This
result indicates that rapid Ru-catalyzed isomerization of the
terminal alkene in 4e led to more energetically stable internal
alkene 4 f prior to hydroesterification (entries 7 and 8). Sub-
strates 4g and 4h, both bearing an exo-alkene moiety, reacted
smoothly, whereas steric effects of substituents adjacent to the
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
4
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
the desired product. Benzyl-
amine was detected in the crude
mixture by 1H NMR analysis,
which indicated that only decar-
bonylation of N-benzylforma-
mide occurred. Despite the opti-
mization of the reaction condi-
tions, the hydrocarbamoylated
product was obtained in only
low yields.
Table 3. Substrate scope of alkenes.[a]
Entry
1
R
Product
Yield
[%][b]
>99
(55:45)
75
(76:24)
Next, intramolecular hydrocar-
bamoylation was tested. Catalyt-
ic hydrocarbamoylation using
[Ru3(CO)12] and L21 afforded the
desired lactams in high yields
(Scheme 3). Compound 6b pre-
dominantly gave the six-mem-
bered lactam, but other sub-
strates including N-alkyls favored
the formation of the five-mem-
bered lactams. Similarly to hy-
droesterification, these results in-
dicated that the regioselectivity
was strongly substrate-depen-
dent. As the starting materials
were easily obtained by the for-
mylation of anilines, Ru-cata-
lyzed hydrocarbamoylation is an
attractive method to synthesize
many oxindole derivatives.
2
3
50
4[c]
39[d]
5
46
83
6
88
(76:24)
7
8[c]
24
69
9
(51:20:29)
66
(39:19:42)
10
Characterization of the Ru–
imidazole complex
[a] Reactions were conducted with 1a (0.40 mmol), 2 (1.5 equiv.), [Ru3(CO)12] (5 mol%), and L21 (15 mol%)
under solvent-free conditions at 1358C for 24 h. [b] Combined isolated yields of the linear and branched iso-
mers. The ratio of isomers (determined by H NMR spectroscopy) is shown in parentheses. [c] The reaction was
1
The combination of Ru and imi-
dazoles was efficient for the cat-
alytic hydroesterification and hy-
drocarbamoylation of alkenes
conducted on a 1.0 mmol scale at 1558C. [d] Mixture of two diastereomers (58:42).
alkenes were clearly observed in the product distribution (en-
tries 9 and 10). Despite steric hindrance, gem-dimethyl-substi-
tuted 4i and trisubstituted alkene 4j afforded products in high
yields (entries 11 and 12). Finally, the Ru–imidazole catalyst was
amenable to the construction of spirocyclic lactones when
cyclic alkenes were used (entries 13 and 14). Use of L17 result-
ed in a better combined yield of 5ka and 5kb (entry 13).
using formates and formamides, respectively. The fact that
both hydroesterification and hydrocarbamoylation did not pro-
ceed without a ligand strongly indicated that the in situ gener-
ated Ru-ligand complex was involved in catalytic cycles. We
tried to identify and characterize an intermediate Ru complex
by mixing [Ru3(CO)12] and 1–3 equivalents of L21 in mesitylene
at 1358C for several hours. However, all attempts resulted in
the formation of an unidentified oily black mixture, and none
of the solid crystallized out of the solution. The long alkyl
chain on the N1 in L21 may have prevented crystallization.
Therefore, L2, in which the dodecyl group of L21 was replaced
by a methyl group, was used for preparation of the Ru com-
plex. After mixing L2 with [Ru3(CO)12] at 1358C, a brown solid
was obtained, which decomposed within 1 day under ambient
conditions and could not be fully characterized. When L20
bearing a benzyl group on the N1 of the imidazole ring was
used, yellow needle-like crystals suitable for X-ray crystallo-
graphic analysis were obtained. The crystallized complex was
unambiguously composed of three Ru atoms and two L20 li-
Application to Ru-catalyzed hydrocarbamoylation
Discovery of ligand acceleration in the Ru-catalyzed hydroes-
terification reaction prompted us to investigate catalytic hydro-
carbamoylation reactions of alkenes using formamides. Metal-
catalyzed hydrocarbamoylation using formamides is an attrac-
tive method for constructing amides from alkenes because of
its synthetic utility and atom economy.[17,18]
Catalytic hydrocarbamoylation was investigated by using the
Ru–imidazole catalyst. Unfortunately, reaction of 4-methoxy-
styrene with N-benzylformamide gave only a trace amount of
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
5
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
conditions for more than 1 year.
Although we also tried to obtain
crystals composed of Ru and
imidazole ligands without the
hydroxy group such as L9, we
did not succeed in obtaining
crystals suitable for structural
studies.
Table 4. Catalytic intramolecular hydroesterification.[a]
Entry
Substrate
Product
Yield
[%][b]
1[c]
2[d]
0
56
(91:9)
75
(90:10)
82
(86:14)
79
(84:16)
72
(82:18)
The ESI–MS spectrum of the
3
Ru–L20
complex
displayed
peak for
a
characteristic
4
5
[Ru3(CO)6(C11H11N2O1)2+H]+, and
C11H11N2O1 seemed to be gener-
ated by the loss of one hydro-
gen atom from L20 (Figure 4).
The isotopic pattern of Ru–L20
was in good agreement with the
calculated one. The same pat-
tern was observed from the re-
action mixture of [Ru3(CO)12] and
L20, which supported retaining
the trinuclear structure in solu-
tion.
6
>99
(90:10)
7
94
(87:13)
8
86
(32:68)
87
9[e]
10[e]
The Ru–L20 complex was ex-
amined for its catalytic ability in
the hydroesterification reaction
(82:18)
(Table 5).
Combination
of
92
(90:10)
11
[Ru3(CO)12] (5 mol%) and L20
(10 mol%) furnished the product
in almost the same yield as the
addition of 15 mol% of L20
(Table 5, entries 1 and 2). This
decrease in the loading of ligand
L20 resulted in an increase in
the linear/branched ratio. The
Ru–L20 complex also promoted
the reaction to afford the de-
sired product in slightly lower
yields. In addition, the pre-
formed catalyst showed a higher
preference for the linear prod-
uct. This result suggests that the
Ru–ligand complex itself or
a species derived from it cata-
lyzes the reaction with a linear
preference, but its catalytic activ-
ity is less than the main catalytic
12
99
75
13[e]
(63:37) [60:40][f]
>99
(72:28) [81:19][f]
14
[a] Reactions were conducted with 4 (100 mg), [Ru3(CO)12] (5 mol%), and L9 (15 mol%) in mesitylene (1.0–
2.0m) at 1358C for 12–24 h. [b] Combined isolated yields of two regioisomers. The ratio of isomers (determined
by H NMR spectroscopy) is shown in parentheses. [c] L9 was not used. [d] L21 was used instead of L9. [e] L17
was used instead of L9. [f] Ratio of diastereomers (shown in brackets) determined by 13C NMR analysis of the
1
isolated mixture of diastereomers. The relative configuration was not determined.
gands (Figure 3), and was similar to the complex previously re-
ported for 2-pyridylmethanol.[19] Each ligand coordinated to
a different Ru atom from imidazole N3 and oxygen in a biden-
tate manner, with oxygen bridging two of the three Ru atoms.
species derived from the in situ formation of [Ru3(CO)12] and
L20. At this juncture, it is unclear whether the real catalytic
species has a mononuclear[8c] or a trinuclear Ru structure.
1
Furthermore, H NMR analysis of a CDCl3 solution of the crys-
Mechanistic studies for catalytic hydroesterification
tals revealed that the peak for the CH2 protons adjacent to the
bridging oxygen atom were split into two doublets, support-
ing the presence of the ring structure. 13C NMR spectroscopy
revealed four carbonyl peaks, also supporting a C2 symmetric
structure. Surprisingly, this crystal was stable under ambient
Although we initially hypothesized the ester exchange mecha-
nism (Scheme 1), we later determined that even imidazoles
without a hydroxy group accelerated Ru-catalyzed hydroesteri-
fication. Therefore, a mechanism other than the ester exchange
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
6
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Scheme 3. Catalytic intramolecular hydrocarbamoylation. The ratio of
1
regioisomers (determined by H NMR spectroscopy) is shown in parentheses.
Figure 4. ESI–MS spectra of the Ru–L20 complex. a) The complete spectrum;
b) the magnified region containing peaks of [Ru3(CO)6(C11H11N2O1)2+H]+;
c) the calculated isotopic pattern of [Ru3(CO)6(C11H11N2O1)2+H]+.
Figure 3. Molecular structure of the Ru–L20 complex with thermal ellipsoids
set at 50% probability.
Table 5. Catalytic
hydroesterification
using
Ru–L20
complex.
Entry
Catalyst
[(mol%)]
Yield
[%][a]
3aa:3ab[b]
1
2
3
[Ru3(CO)12] (5)+L20 (15)
[Ru3(CO)12] (5)+L20 (10)
Ru–L20 (5)
73
72
58
37:63
58:42
71:29
[a] Combined isolated yields of 3aa and 3ab. [b] Determined by 1H NMR
spectroscopy.
mechanism must be operative. On the basis of our results and
the literature precedent,[6,7a,8c,16a] we propose the mechanism
shown in Scheme 4. In this mechanism, the catalytic activity of
Ru is enhanced by the imidazole ligands. In path i, ruthenium
Scheme 4. Proposed reaction mechanism. L=ligand.
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
7
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
hydride A undergoes alkene insertion. In path ii, initial decar-
bonylation of A forms ruthenium hydride B, which undergoes
alkene insertion followed by recarbonylation. Similar decarbon-
ylation–recarbonylation processes were proposed for hydroes-
terification reactions using formate.[7a,8b,c] Because CO insertion
could occur at the metalÀO[20] or metalÀC bond, reductive
elimination finally gives the product.
Table 7. Ru-catalyzed
hydroesterification
using
BnOH.
Entry
Catalyst
Yield
[%][a]
3aa:3ab[b]
[(mol%)]
Experimental differentiation between paths i and ii in
Scheme 4 was difficult, but we could confirm the presence of
the decarbonylation–recarbonylation pathway through label-
ing experiments. Benzyl [13C]formate (1a-13C) was prepared
from [13C]formic acid and used in the hydroesterification reac-
tion to verify 13C incorporation in the product (Table 6). The re-
1
2
[Ru3(CO)12] (10)+L20 (20)
Ru–L20 (10)
34
24
64:36
81:19
[a] Combined isolated yield of 3aa and 3ab. [b] Determined by 1H NMR
spectroscopy.
tion was conducted using 10 mol% of the Ru–L20 complex
(0.80 equiv. of CO source), the product was obtained in 24%
yield. This decrease in yield was reasonable for the reduction
in the total amount of CO source (entry 2). These results clearly
indicate the presence of decarbonylation and recarbonylation
steps in the catalytic cycle.
Table 6. Ru-catalyzed
hydroesterification
using
1a-13C.
Next, a labeling study using benzyl formate-d1 (1a-D) in the
Ru-catalyzed hydroesterification was conducted. Deuterium-la-
beled formate would be effective for obtaining mechanistic in-
sight into the rate-determining step and reversibility of the
alkene insertion step. A similar mechanistic study using deute-
rium-labeled formate was conducted by Floreancig and Wang
for the Ru-catalyzed tandem hydroesterification and lactoniza-
tion reaction.[16] Therefore, 1a-D was prepared according to
the literature precedent[21] and used for deuterium-labeling
study (Scheme 5). The yield of the product obtained using
Entry Catalyst
[(mol%)]
Yield 3aa-13C:3ab-13C[b] 13C incorporation [%][c]
[%][a]
3aa-13C/3ab-13C
1[d]
[Ru3(CO)12] (5) 80
L20 (15)
38:62
38:62
72:28
–
2
[Ru3(CO)12] (5) 72
53/57
L20 (15)
3
Ru-L20 (5)
56
61/70
[a] Combined isolated yields of 3aa-13C to 3ab-13C. [b] Determined by
1H NMR spectroscopy. [c] Calculated by 13C NMR spectroscopy. [d] 1a (not
13C-labeled) was used instead of 1a-13C.
action using 1a-13C proceeded similarly to the reaction using
unlabeled 1a (Table 6, entries 1 and 2). The 13C incorporation
in products 3aa and 3ab was approximately 50–60%. Partial
incorporation of 13C indicated the existence of an alternative
recarbonylation pathway for the incorporation of unlabeled
carbonyl groups originating from [Ru3(CO)12]. This result sug-
gests that the labeled formate undergoes initial decarbonyla-
tion and the 13CO formed competes with CO from [Ru3(CO)12]
in the recarbonylation step. As 13CO from 1a-13C corresponds
to 1.5 equivalents and CO from [Ru3(CO)12] corresponds to
0.60 equivalents, less than a statistical amount of 13C resulted
in incorporation (entry 2). Therefore, it can be assumed that
13CO from 1a-13C did not well accumulate at the beginning of
the reaction and thus CO from [Ru3(CO)12] was preferentially in-
corporated into the products. Conducting the reaction with
the isolated Ru–imidazole complex, a theoretical source of
0.40 equivalents of 12C carbonyl groups demonstrated a reason-
able increase in 13C incorporation (entry 3). In this case, the re-
gioselectivity was reversed with a strong preference for the
linear product.
Scheme 5. Ru-catalyzed hydroesterification using 1a-D.
1a-D was similar to that for unlabeled 1a (Scheme 5 vs.
Table 5, entry 1). Deuterium incorporation at carbons a, b, and
d but not at c suggests that alkene insertion for the branched
isomer is reversible through b-elimination, and alkene insertion
for the linear isomer is irreversible. It should be noted that the
deuterium incorporation for each regioisomer is low, suggest-
ing that the species having RuÀH, which was generated by the
b-elimination process, would undergo alkene insertion prefer-
entially over the species having RuÀD because of isotope
effect. Hydrogen derived from sources other than the formyl
CÀH might have also been incorporated, although these sour-
ces have not been identified so far.[22]
The decarbonylation–recarbonylation pathway was further
confirmed by the reaction using benzyl alcohol instead of
benzyl formate (Table 7). This reaction afforded the desired
product in 34% yield using 10 mol% of [Ru3(CO)12], i.e.,
1.2 equivalents of the CO source (Table 7, entry 1). If the reac-
Although the mechanism shown in Scheme 4 does not
invoke the initially hypothesized ester exchange mechanism
for imidazoles with a hydroxy group, this process cannot be
conclusively ruled out. In fact, ester exchange of 1a with L2 in
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
8
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
1358C (bath temperature) and stirred for 24 h. The reaction mix-
ture was cooled to RT and diluted with EtOAc, washed with H2O
three times, dried over Na2SO4, filtered, and concentrated. The resi-
due was purified by column chromatography on silica gel (hexane/
EtOAc 5:1) to afford a mixture of the desired linear and branched
product.
Representative procedure for intramolecular hydroesterifica-
tion (Table 4, entry 3)
Compounds 4a (100 mg, 0.62 mmol), L9 (15.9 mg, 0.093 mmol,
15 mol%), and mesitylene (0.30 mL) were added to a 2 mL vial
equipped with a silicon septum cap under a flow of Ar. [Ru3(CO)12]
(19.8 mg, 0.031 mmol, 5 mol%) was added to the vial and then
sealed by a new silicon septum cap. The mixture was warmed to
1358C (bath temperature) and stirred for 24 h. The reaction mix-
ture was cooled to RT and directly purified by preparative TLC on
silica gel (hexanes/EtOAc 4:1) to afford six-membered lactone 5aa
(67.5 mg, 0.42 mmol, 68%) as a colorless oil and seven-membered
lactone 5ab (7.8 mg, 0.048 mmol, 8%) as a colorless oil.
Scheme 6. Reactions used to confirm the ester exchange steps.
mesitylene at 1358C occurred in 3 h to produce ester 8
(Scheme 6). In addition, the reaction of a mixture of 9a and
9b with BnOH in the presence of [Ru3(CO)12] gave the corre-
sponding benzyl esters. Thus, the two ester exchange steps
shown in Scheme 1 occur under the hydroesterification condi-
tions. As further investigations are needed to conclusively
identify the mechanism for hydroesterification, at this juncture,
we cannot exclude the presence of the ester exchange mecha-
nism if ligands bearing a hydroxy group are used.
Representative procedure for intramolecular hydrocarba-
moylation (Scheme 3, first reaction)
Compounds 6a (100 mg, 0.68 mmol), L21 (27.1 mg, 0.10 mmol,
15 mol%), and mesitylene (0.30 mL) were added to a 2 mL vial
equipped with a silicon septum cap under a flow of Ar. [Ru3(CO)12]
(21.7 mg, 0.034 mmol, 5 mol%) was added to the vial and then
sealed by a new silicon septum cap. The mixture was warmed to
1358C (bath temperature) and stirred for 23 h. The reaction mix-
ture was cooled to RT and directly purified by column chromatog-
raphy on silica gel (10 g, hexanes/EtOAc 4:1) to afford five-mem-
bered lactam 7aa (72.4 mg, 0.49 mmol, 72%) as a white solid and
six-membered lactam 7ab (8.1 mg, 0.055 mmol, 8%) as a white
needle-like solid.
Conclusions
Imidazole derivatives were identified as highly efficient ligands
for drastically accelerating the [Ru3(CO)12]-catalyzed hydroester-
ification of alkenes with formates. 2-Hydroxymethylimidazoles
were especially effective for intermolecular hydroesterification,
whereas simpler alkylimidazoles were suitable for intramolecu-
lar hydroesterification. This novel Ru-imidazole catalyst system
promotes the hydroesterification of various formates and al-
kenes to afford a series of one-carbon homologated esters.
Modification of the ligand structure improved the regioselec-
tivity. Moreover, Ru-catalyzed intramolecular hydrocarbamoyla-
tion was effective for the construction of lactams. A complex
obtained from [Ru3(CO)12] and 2-hydroxymethylimidazole was
unambiguously characterized by NMR, ESI–MS, and X-ray crys-
tallographic analyses, and was successfully used for the hydro-
esterification reaction. The reactions promoted by the Ru–imi-
dazole catalyst system proceed by a decarbonylation–recarbo-
nylation pathway, which was confirmed for the first time by
a series of mechanistic studies. Further applications of this
novel Ru–imidazole catalyst are underway in our laboratory.
Acknowledgements
We thank Dr. Yoshinobu Ishikawa (University of Shizuoka) for his
assistance with the X-ray crystallographic analysis. This work was
partly supported by Daiichi-Sankyo Co. Ltd., the Uehara Memori-
al Foundation, JSPS Grant-in-Aid for Young Scientists (B,
23790016), and Platform for Drug Discovery, Informatics, and
Structural Life Science from the Ministry of Education, Culture,
Sports, Science and Technology, Japan.
Keywords: alkenes · carbonylation · ligand design · reaction
mechanisms · ruthenium
[1] a) J. Mulzer in Comprehensive Organic Synthesis, Vol. 6 (Eds.: B. M. Trost,
I. Fleming), Pergamon, Oxford, 1991, pp. 323–380; b) A. R. M. Yahya,
W. A. Anderson, M. Moo-Young, Enzyme Microb. Technol. 1998, 23, 438–
450.
[2] G. Kiss, Chem. Rev. 2001, 101, 3435–3456.
[3] W. Reppe, H. Kroper, H. J. Pistor, H. Schlenck, Justus Liebigs Ann. Chem.
1953, 582, 38–71.
[4] P. Isnard, B. Denise, R. P. A. Sneeden, J. Organomet. Chem. 1983, 256,
135–139.
[5] For selected early examples of catalytic hydroesterification, see: a) M.
Mlekuz, F. Joo, H. Alper, Organometallics 1987, 6, 1591–1593; b) W.
Ueda, T. Yokoyama, Y. Morikawa, Y. Moro-oka, T. Ikawa, J. Mol. Catal.
Experimental Section
General procedure for intermolecular hydroesterification
(Tables 2 and 3)
Formate (0.40 mmol), L21 (16.0 mg, 0.060 mmol, 15 mol%), and
[Ru3(CO)12] (12.8 mg, 0.020 mmol, 5 mol%) were added to a 2 mL
vial equipped with a silicon septum cap under a flow of Ar. The
alkene (0.60 mmol, 1.5 equiv.) was added to the vial and then
sealed by a new silicon septum cap. The mixture was warmed to
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
9
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
1988, 44, 197–200; c) T. Kondo, S. Yoshii, Y. Tsuji, Y. Watanabe, J. Mol.
Catal. 1989, 50, 31–38; d) W. Keim, J. Becker, J. Mol. Catal. 1989, 54,
95–101; e) E. M. Nahmed, G. Jenner, J. Mol. Catal. 1990, 59, L15–L19;
f) C. Legrand, Y. Castanet, A. Mortreux, F. Petit, J. Chem. Soc. Chem.
Commun. 1994, 1173–1174; g) Y. Suzuki, H. Katoh, Y. Ishii, M. Hidai, J.
Mol. Catal. A 1995, 95, 129–133; h) S. Fabre, P. Kalck, G. Lavigne, Angew.
Chem. Int. Ed. Engl. 1997, 36, 1092–1095; Angew. Chem. 1997, 109,
1167–1169; i) T. Kondo, T. Okada, T. Mitsudo, Organometallics 1999, 18,
4123–4127.
[16] For lactone synthesis including intermolecular hydroesterification and
subsequent cyclization, see: a) L. Wang, P. E. Floreancig, Org. Lett. 2004,
6, 4207–4210. See also: b) H. Wang, B. Dong, Y. Wang, J. Li, Y. Shi, Org.
Lett. 2014, 16, 186–189.
[17] a) S. Ko, H. Han, S. Chang, Org. Lett. 2003, 5, 2687–2690; b) D. C. D.
Nath, C. M. Fellows, T. Kobayashi, T. Hayashi, Aust. J. Chem. 2006, 59,
218–224; c) N. Armanino, E. M. Carreira, J. Am. Chem. Soc. 2013, 135,
6814–6817; d) P. A. Donets, N. Cramer, J. Am. Chem. Soc. 2013, 135,
11772–11775; e) K. S. Kanyiva, F. Lçbermann, Y. Nakao, T. Hiyama, Tetra-
hedron Lett. 2009, 50, 3463–3466; f) Y. Miyazaki, Y. Yamada, Y. Nakao, T.
Hiyama, Chem. Lett. 2012, 41, 298–300.
[18] During the preparation of this paper, Chang and co-workers reported
their work on a similar catalytic hydrocarbamoylation of alkenes using
[Ru3(CO)12] and a quaternary ammonium salt. B. Li, Y. Park, S. Chang, J.
Am. Chem. Soc. 2014, 136, 1125–1131.
[19] a) S. Ko, C. Lee, M.-G. Choi, Y. Na, S. Chang, J. Org. Chem. 2003, 68,
1607–1610; b) D. J. Darensbourg, B. Fontal, S. S. Chojnacki, K. K. Klaus-
meyer, J. H. Reibenspies, Inorg. Chem. 1994, 33, 3526–3532; c) J. A. van
Doorn, P. W. N. M. van Leeuwen, J. Organomet. Chem. 1981, 222, 299–
309; d) CCDC- 1031455 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif.
[6] S. Ko, Y. Na, S. Chang, J. Am. Chem. Soc. 2002, 124, 750–751.
[7] a) Y. Katafuchi, T. Fujihara, T. Iwai, J. Terao, Y. Tsuji, Adv. Synth. Catal.
2011, 353, 475–482; b) I. Fleischer, R. Jennerjahn, D. Cozzula, R. Jack-
stell, R. Franke, M. Beller, ChemSusChem 2013, 6, 417–420.
[8] a) H. Konishi, T. Ueda, T. Muto, K. Manabe, Org. Lett. 2012, 14, 4722–
4725; b) H. Konishi, K. Manabe, Synlett 2014, 25, 1971–1986; c) After
our initial report, Profir, Beller, and Fleischer. also reported Ru-catalyzed
hydroesterification using an imidazole-based ligand: I. Profir, M. Beller, I.
Fleischer, Org. Biomol. Chem. 2014, 12, 6972–6976.
[9] For a review of metal-catalyzed carbonylation without the use of CO,
see: a) T. Morimoto, K. Kakiuchi, Angew. Chem. Int. Ed. 2004, 43, 5580–
5588; Angew. Chem. 2004, 116, 5698–5706; b) L. Wu, Q. Liu, R. Jackstell,
M. Beller, Angew. Chem. Int. Ed. 2014, 53, 6310–6320; Angew. Chem.
2014, 126, 6426–6436.
[20] a) H. E. Bryndza, Organometallics 1985, 4, 1686–1687; b) I. Tꢂth, C. J.
Elsevier, J. Chem. Soc. Chem. Commun. 1993, 529–531; c) Y. Hu, J. Liu, Z.
Lꢃ, X. Luo, H. Zhang, Y. Lan, A. Lei, J. Am. Chem. Soc. 2010, 132, 3153–
3158.
[21] Z. Wu, S. T. Nguyen, R. H. Grubbs, J. W. Ziller, J. Am. Chem. Soc. 1995,
117, 5503–5511.
[10] For selected reviews, see: a) P. Anastas, N. Eghbali, Chem. Soc. Rev.
2010, 39, 301–312; b) R. A. Sheldon, Chem. Soc. Rev. 2012, 41, 1437–
1451.
[11] For a use of a stoichiometric amount of L1 in hydroesterification to pro-
duce pyridylmethyl esters, see: S. Inoue, K. Yokota, H. Tatamidani, Y. Fu-
kumoto, N. Chatani, Org. Lett. 2006, 8, 2519–2522.
[22] a) H/D exchange reaction catalyzed by Ru hydride was reported, see:
S. K. S. Tse, P. Xue, Z. Lin, G. Jia, Adv. Synth. Catal. 2010, 352, 1512–
1522; b) For a recent review of metal-catalyzed H/D exchange, see: Y.
Sawama, Y. Monguchi, H. Sajiki, Synlett 2012, 23, 959–972.
[12] The use of [RuCl2(cod)] (cod=1,5-cyclooctadiene) or [RuClH(CO)(PPh3)3]
instead of [Ru3(CO)12] afforded the product in less than 1% yield.
[13] For decarbonylation of formates by Ru–phosphine, see: G. Jenner, E. M.
Nahmed, H. Leismann, J. Organomet. Chem. 1990, 387, 315–321.
[14] a) K. Dallmann, R. Buffon, J. Mol. Catal. A 2001, 172, 81–87; b) H. B.
Ammar, J. L. Nꢁtre, M. Salem, M. T. Kaddachi, L. Toupet, J.-L. Renaud, C.
Bruneau, P. H. Dixneuf, Eur. J. Inorg. Chem. 2003, 4055–4064.
[15] N. Armanino, M. Lafrance, E. M. Carreira, Org. Lett. 2014, 16, 572–575.
Received: December 4, 2014
Revised: December 19, 2014
Published online on && &&, 0000
&
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
10
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

FULL PAPERS
H. Konishi, T. Muto, T. Ueda, Y. Yamada,
M. Yamaguchi, K. Manabe*
&& – &&
Imidazole Derivatives as Accelerators
for Ruthenium-Catalyzed
Effective imidazole assistant:
both alkenes and formates. The Ru–
imidazole complex also catalyzes the in-
tramolecular hydrocarbamoylation of al-
kenes.
Hydroesterification and
[Ru3(CO)12]-catalyzed hydroesterification
of alkenes by using formates is drastical-
ly accelerated by imidazole derivatives
and exhibits a broad substrate scope for
Hydrocarbamoylation of Alkenes:
Extensive Ligand Screening and
Mechanistic Study
ChemCatChem 0000, 00, 0 – 0
www.chemcatchem.org
11
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ