Amide synthesis from alcohols and amines catalyzed by ruthenium N-Heterocyclic carbene complexes

Article abstract of DOI:10.1002/chem.201000569

The direct synthesis of amides from alcohols and amines is described with the simultaneous liberation of dihydrogen. The reaction does not require any stoichiometric additives or hydrogen acceptors and is catalyzed by ruthenium N-heterocyclic carbene complexes. Three different catalyst systems are presented that all employ 1,3-diisopropylimidazol-2-ylidene (IiPr) as the carbene ligand. In addition, potassium iert-butoxide and a tricycloalkylphosphine are required for the amidation to proceed. In the first system, the active catalyst is generated in situ from [RuCl₂(cod)] (cod = 1,5-cyclooctadiene), 1,3-diisopropylimidazolium chloride, tricyclopentylphosphonium tetrafluoroborate, and base. The second system uses the complex [RuCl ₂(IiPr)(p-cymene)] together with tricyclohexylphosphine and base, whereas the third system employs the Hoveyda-Grubbs lst-generation metathesis catalyst together with 1,3-diisopropylimidazolium chloride and base. A range of different primary alcohols and amines have been coupled in the presence of the three catalyst systems to afford the corresponding amides in moderate to excellent yields. The best results are obtained with sterically unhindered alcohols and amines. The three catalyst systems do not show any significant differences in reactivity, which indicates that the same catalytically active species is operating. The reaction is believed to proceed by initial dehydrogenation of the primary alcohol to the aldehyde that stays coordinated to ruthenium and is not released into the reaction mixture. Addition of the amine forms the hemiaminal that undergoes dehydrogenation to the amide. A catalytic cycle is proposed with the {(I(Pr)Ru^II} species as the catalytically active components.

Full text of DOI:10.1002/chem.201000569

DOI: 10.1002/chem.201000569
Amide Synthesis from Alcohols and Amines Catalyzed by Ruthenium
N-Heterocyclic Carbene Complexes
Johan Hygum Dam, Gyorgyi Osztrovszky, Lars Ulrik Nordstrøm, and Robert Madsen*[a]
Abstract: The direct synthesis of
amides from alcohols and amines is de-
scribed with the simultaneous libera-
tion of dihydrogen. The reaction does
not require any stoichiometric addi-
tives or hydrogen acceptors and is cata-
lyzed by ruthenium N-heterocyclic car-
bene complexes. Three different cata-
lyst systems are presented that all
employ 1,3-diisopropylimidazol-2-yli-
dene (IiPr) as the carbene ligand. In
addition, potassium tert-butoxide and a
tricycloalkylphosphine are required for
the amidation to proceed. In the first
system, the active catalyst is generated
um tetrafluoroborate, and base. The
second system uses the complex
results are obtained with sterically un-
hindered alcohols and amines. The
three catalyst systems do not show any
significant differences in reactivity,
which indicates that the same catalyti-
cally active species is operating. The re-
action is believed to proceed by initial
dehydrogenation of the primary alco-
hol to the aldehyde that stays coordi-
nated to ruthenium and is not released
into the reaction mixture. Addition of
the amine forms the hemiaminal that
undergoes dehydrogenation to the
amide. A catalytic cycle is proposed
with the {(IiPr)RuII} species as the cat-
alytically active components.
[RuCl
2ACHUTGTNERNNUG(IiPr)ACHTGNUTRENNUNG
tricyclohexylphosphine
whereas the third system employs the
Hoveyda–Grubbs 1st-generation meta-
thesis catalyst together with 1,3-diiso-
propylimidazolium chloride and base.
A range of different primary alcohols
and amines have been coupled in the
presence of the three catalyst systems
to afford the corresponding amides in
moderate to excellent yields. The best
Keywords: alcohols
· amides ·
in situ from [RuCl2ACHTUNGTRENNUNG(cod)] (cod=1,5-cy-
clooctadiene), 1,3-diisopropylimidazoli-
um chloride, tricyclopentylphosphoni-
carbene ligands · dehydrogenation ·
ruthenium
Introduction
large activation energy due to the formation of the corre-
sponding ammonium salt and this method generally requires
a temperature above 1608C.[2] The temperature can be sig-
nificantly lowered by catalyzing the dehydration with spe-
cially designed areneboronic acids[3] or heterogeneous silica
catalysts[4] if water at the same time is removed irreversibly.
The most common methods for amide synthesis employ acti-
vated derivatives of the carboxylic acid, such as the chloride
and the anhydride.[2] The activated derivatives may also be
generated in situ by employing stoichiometric coupling re-
agents, such as carbodiimides, uronium, and phosphonium
salts,[5] for which the latter two are the methods of choice in
peptide synthesis. Other general procedures for amide syn-
thesis include the Beckman rearrangement,[6] Staudinger li-
gations,[7] oxidative amidation of aldehydes,[8] coupling of a-
ketoacids and hydroxylamines,[9] and amidation of ketones
and thioacids with azides.[10] More recently, a number cata-
lytic procedures have been developed including amidation—
hydrolysis of gem-dihaloolefins,[11] redox rearrangement of
a-functionalized aldehydes,[12] and aminocarbonylation of
aryl halides and terminal alkynes.[13]
The amide is one of the most prevalent linkages in organic
chemistry. It is the key functional group in peptides and a
number of polymers and is also found in many pharmaceuti-
cals and natural products.[1] The synthesis of amides has
been the subject of intense studies and numerous methods
have been developed.[2] However, cost effective, high-yield-
ing and waste-free procedures with a broad substrate scope
are still in high demand. The direct synthesis of amides by
thermal dehydration of carboxylic acids and amines has a
[a] Dr. J. H. Dam, G. Osztrovszky, Dr. L. U. Nordstrøm,
Prof. Dr. R. Madsen
Department of Chemistry, Building 201
Technical University of Denmark
2800 Kgs. Lyngby (Denmark)
Fax : (+45)4593-3968
E-mail: rm@kemi.dtu.dk
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201000569.
6820
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6820 – 6827

FULL PAPER
Table 1. Amidation with catalysts generated in situ.
Very recently, amide synthesis has become possible by the
direct metal-catalyzed coupling of primary alcohols and
amines with the concomitant extrusion of dihydrogen
(Scheme 1). The reaction presumably occurs by initial dehy-
Entry
R
X
Ligand
PPh3
Yield [%][a]
1
2
3
4
5
6
7
8
Mes
Mes
Mes
Mes
Mes
Mes
Mes
Mes
Me
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
21
26
26
22
27
24
24
12
53
P
P
ACHTUNGTRENNUNG
A
PtBu3
PCy3
O=PPh3
AsPh3
pyridine
PCy3
Scheme 1. Dehydrogenative amide formation from primary alcohols and
amines.
9
(MeO)2PO2
10
11
12
13
14
iPr
Cy
tBu
iPr
iPr
Cl
PCy3
PCy3
PCy3
PCyp3
PCyp3·HBF4
92
84
68
BF4
BF4
Cl
drogenation of the alcohol to the aldehyde followed by
hemiaminal formation with the amine and subsequent dehy-
drogenation to the amide. The amidation has been achieved
both in the presence[14,15] and absence[16–18] of hydrogen scav-
engers. The latter protocol is the most attractive in which no
stoichiometric additives are necessary and dihydrogen is
produced as the only byproduct. To date, three different cat-
alyst systems have been reported for this atom-economical
amidation procedure for which two are homogeneous proto-
cols and the latter a heterogeneous method. The first system
was presented by Milstein et al. in 2007 for which a rutheni-
um complex with a PNN-type pincer ligand was shown to be
an effective catalyst for the coupling of primary alcohols
and amines with the liberation of dihydrogen.[16] The follow-
ing year our laboratory showed that the same transforma-
tion could be performed with an in situ generated ruthenium
N-heterocyclic carbene (NHC) catalyst.[17,19] In 2009, Shimi-
zu et al. achieved the dehydrogenative amide synthesis with
a silver cluster supported on g-alumina as the catalyst.[18] Of
these three systems the in situ generated ruthenium carbene
catalyst is easily modified and can be carried out with com-
mercially available reagents.
98
Cl
92[b]
[a] GC yield. [b] With 20% of KOtBu.
ed for further studies (Table 1, entries 1–8). The influence of
the substitutents on the N-heterocyclic carbene was then in-
vestigated in detail. These substituents had a pronounced
impact on the amidation and the isopropyl group was found
to give the highest yield (entries 9–12). A number of more
substituted imidazolium salts gave less than 25% yield
under the same conditions.[20] Carbenes with a saturated
backbone, that is, imidazolin-2-ylidenes, gave significantly
lower yields than carbenes with an unsaturated backbone.[17]
Potassium tert-butoxide was selected as the base for generat-
ing the carbene since it is easy to handle. Similar yields were
obtained with potassium hexamethyldisilazide, whereas the
use of cesium carbonate resulted in lower yields. The pur-
pose of the base is not only to deprotonate the imidazolium
salt, but also to promote the amide formation. Various
amounts of base were investigated and the optimum amount
was found to be three times the amount of the imidazolium
salt. With 1,3-diisopropylimidazol-2-ylidene as the carbene
of choice, the phosphine ligand was investigated again. In
this case, tricyclopentylphosphine (PCyp3) gave a slight im-
provement over PCy3. The improvement was not only mea-
sured in the yield at the end of the reaction, but also after
3 h when PCyp3 showed 67% conversion and PCy3 only
56% (entries 10 and 13). However, PCyp3 is a liquid and sig-
nificantly less stable than the tricyclohexyl congener. There-
fore, the corresponding crystalline HBF4 salt[21] was em-
ployed at the expense of additional base (entry 14). The iso-
lated yields from the experiments in entries 13 and 14 were
the same and the catalyst system in entry 14 was selected for
general use and denoted catalyst A.
Herein, we report a full account on our studies of rutheni-
um N-heterocyclic carbene catalysts in the dehydrogenative
amidation from primary alcohols and amines. We demon-
strate that the reaction can be achieved with three different
(pre)catalysts and provide further support for the catalyti-
cally active species.
Results and Discussion
Catalyst development: 2-Phenylethanol and benzylamine
were selected as test substrates for optimizing the amidation
procedure. Initial experiments revealed that the reaction
could be achieved with a ruthenium(II) precursor in the
presence of an in situ generated N-heterocyclic carbene
(Table 1). To prevent rapid deactivation of the catalyst, it
was also necessary to add an additional ligand. A range of
phosphine ligands and other ligands could be used for this
purpose for which PCy3 gave the best result and was select-
Since the catalytically active species in this reaction may
be a ruthenium(II)chloride N-heterocyclic carbene complex
it would be of interest to study the reaction with a more
well-defined complex. This may lead to a new catalyst
system and a better understanding of the mechanism. At-
Chem. Eur. J. 2010, 16, 6820 – 6827
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6821

R. Madsen et al.
Table 2. Amidation with [RuACTHNUGRTENNUG(NHC)ACHTUTGNREN(NUGN p-cymene)] complexes.
tempts to isolate a carbene complex from the reaction be-
tween 1,3-diisopropylimidazolium chloride, [RuCl2A(cod)],
CTHUNGTRENNUNG
phosphine, and base were not successful and the in situ gen-
erated carbene complex appears to be very sensitive. In-
stead, we turned our attention to the known p-cymene com-
plexes of ruthenium(II)chloride and N-heterocyclic car-
benes.[22] These are stable and coordinatively saturated com-
plexes that have been used for hydrogenation and cyclopro-
panation of olefins.[22] It is known that the p-cymene ligand
is released at about 858C[23] and with the amidation being
performed in refluxing toluene these complexes appear well
suited as catalyst precursors. Traditionally, the p-cymene
complexes have been prepared by transfer of the free N-het-
Entry Complex Phosphine Yield [%] (3 h)[a] Yield [%] (24 h)[a]
[b]
1
–
–
1
1
2
2
3
3
–
–
PCy3
56
67
65
53
61
56
22
29
56
63
92
98
[b]
2
3
PCyp3
PCy3
95
4
5
PCyp3
PCy3
100
97
6
7
PCyp3
PCy3
91
50
8
9
10
PCyp3
PCy3
PCyp3
44
90
87
[c]
[c]
erocyclic carbene to [RuCl2ACTHNUTRGNEUNG
(p-cymene)]2.[24] More recently,
the carbene transfer has become possible by reaction of 1,3-
dialkylimidazolium chlorides with silver oxide in dichloro-
methane.[25] By this method the corresponding silver carbene
[a] GC yield. [b] Generated in situ from [RuCl2ACTHNUTRGENUN(G cod)]2 and 1,3-diisopro-
pylimidazolium chloride (catalyst A). [c] Generated in situ from [RuI
cymene)]2 and 1,3-dicyclohexylimidazolium chloride.
2ACHTUNGTRENNUNG(p-
is generated and transmetallated in situ with [RuCl2ACTHNUTRGNE(UNG p-
cymene)]2, which makes it unnecessary to isolate the free
carbene. In this way, complexes 1 and 2 were generated in
excellent yield and isolated by flash chromatography
(Scheme 2). The structure of 2 has previously been con-
firmed by X-ray crystallography.[24c] Except for the two dif-
clic carbene ruthenium(II)chloride species is produced
under the in situ conditions. Diiodide complex 3, on the
other hand, was less reactive than dichlorides 1 and 2 and
more byproducts were formed with this complex (entries 7
and 8). The results with diiodide 3 did not improve by
adding 10% of lithium chloride or tetraethylammonium
chloride to the reaction. However, when the amidation was
performed with an in situ generated catalyst from [RuI2ACHTUNGTRENNUNG(p-
cymene)]2, 1,3-dicyclohexylimidazolium chloride, phosphine,
and base almost the same yields were observed as with
1
ferent alkyl groups, the H and 13C NMR spectroscopic data
for 1 and 2 are very similar with the carbene carbon atom in
both cases located at d=171 ppm in the 13C NMR spectrum.
To probe the influence of the halide on ruthenium, the cor-
responding diiodide complex 3 was also prepared. In this
case, only a 56% yield of 3 was obtained since the carbene
transfer between 1,3-dicyclohexylimidazolium chloride and
[RuCl2ACHTNUTRGNE(NUG cod)] as the ruthenium precursor (entries 9 and 10).
[RuI
(p-cymene)]2 gave a mixture of dichloride 2 and diio-
We believe the catalytically active complex in this case is
mainly a ruthenium(II)chloride species and not the corre-
sponding iodide complex. No reaction occurred when the
amidation was attempted with 5% of 1 and 10% of silver
triflate in the presence of phosphine and base. Based on
these studies we selected complex 1 together with PCy3 for
general use and denoted this system catalyst B.
dide 3 that were separated by preparative TLC.
In 2001 Grubbs et al. showed that the Grubbs 2nd-genera-
tion metathesis catalyst reacts with dihydrogen to remove
the benzylidene ligand, but not the N-heterocyclic carbene
ligand.[26] This observation prompted us to investigate olefin
metathesis catalysts[27] since the liberated dihydrogen in the
amidation may serve to activate the metathesis catalysts for
this transformation. Indeed, reaction of 2-phenylethanol and
benzylamine with Grubbs 2nd-generation catalyst and base
produced the amide in 49% yield after 24 h (Table 3,
Scheme 2. Synthesis of NHC ruthenium cymene complexes.
Complexes 1–3 were tested in the amidation with 2-phe-
nylethanol and benzylamine, and the yield was measured
after both 3 and 24 h (Table 2). Again, the reaction required
a base for the amidation to proceed. A phosphine was also
required to obtain a high yield of the amide. Without phos-
entry 1). This is a lower yield than that achieved in TaACHTUNGTRENNUNGbles 1
and 2, but the saturated N-heterocyclic carbene in Grubbs
2nd-generation catalyst is not the optimum ligand for the
amidation. A higher yield was obtained with Hoveyda–
Grubbs 2nd-generation catalyst and this did not change by
adding PCy3 to the reaction (Table 3, entry 2). Interestingly,
the Grubbs catalyst with the less sterically demanding o-
tolyl group[28] gave a good yield of the amide (entry 3).
To illustrate the influence of the N-heterocyclic carbene
Grubbs 1st-generation and Hoveyda–Grubbs 1st-generation
catalysts were also investigated. These two complexes do
phine, less than 70% of the amide was observed after 24 h.
[21]
The phosphine salt PCyp3·HBF4
was less effective with
complexes 1 and 2 and afforded below 70% yield of the
amide after 24 h. However, with added PCy3 and PCyp3
complexes 1 and 2 performed very well in the amidation
(Table 2, entries 3–6) and gave results after 3 and 24 h which
were very similar to the yields from the in situ generated
catalyst (entries 1 and 2). This confirms that an N-heterocy-
6822
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6820 – 6827

Amide Synthesis from Alcohols and Amines
Table 3. Amidation with metathesis catalysts.
FULL PAPER
Entry
1
Metathesis catalyst
Yield [%] (3 h)[a]
Yield [%] (24 h)[a]
29
49
ed in situ (entries 6 and 12). Based on the results in Table 3,
Hoveyda–Grubbs 1st-generation catalyst was selected as the
metathesis catalyst for the amidation in the presence of 1,3-
diisopropylimidazolium chloride and base (catalyst C).
2
48
65
3
63
92
Substrate scope: With three optimized catalysts in hand, the
substrate scope and limitations could now be more thor-
oughly explored. Equimolar amounts of various primary al-
cohols and amines were reacted with catalysts A, B, and C
to afford the corresponding amides (Table 4). Sterically un-
hindered alcohols reacted with primary amines to give the
secondary amide in high yields (Table 4, entries 1–3). Benzyl
alcohol furnished the corresponding benzamide (entry 4),
whereas the aryl chloride in entry 5 afforded the amide
without concomitant dechlorination. Hex-5-en-1-ol, on the
other hand, gave exclusively the hexanamide with all three
catalysts in which the olefin had been reduced with the li-
berated dihydrogen (entry 6). N-benzylethanolamine under-
went coupling with benzylamine in high yield, which illus-
trates that the amidation is selective for a primary amine
over a secondary amine. Optically pure 1-phenylethylamine
participated in the amidation without any sign of epimeriza-
tion (entry 8). The same was observed with optically pure
N-benzyl-l-prolinol (entry 9), which is noteworthy since the
reaction goes through the corresponding aldehyde. Prolinol
gave a lower yield than the other primary alcohols and was
not completely consumed in the amidation, which may re-
flect the slightly higher steric demand around this alcohol.
The reaction could also be performed in an intramolecular
fashion to afford both five- and seven-membered lactams
(entries 10 and 11). On the contrary, aniline and secondary
amines did not react with 2-phenylethanol in refluxing tolu-
ene. In these cases, the amidation was carried out in reflux-
ing mesitylene, which gave a moderate yield with aniline
(entry 12) and a good yield with the secondary amine
(entry 13). In the last two cases, self-condensation of the al-
cohol into the corresponding ester was observed as a by-
product, whereas the other examples in Table 4 did not
reveal any single compound as a major byproduct. Several
other alcohols and amines reacted very poorly or not at all
in refluxing mesitylene. N-Boc-protected ethanolamine, 1-
phenylethane-1,2-diol, 2-pyridineethanol, and 2-(4-bromo-
phenyl)ethanol only gave trace amounts of the amide in the
reaction with benzylamine. Several derivatives of glycine[32]
also failed to give more than trace amounts of the amide in
the reaction with 2-phenylethanol. Compared with the re-
sults in Table 4, these examples illustrate that the amidation
4
5
6
48
76[b]
76[c]
71
100[b]
96[c]
7
8
9
41
72[b]
76[c]
60
97[b]
100[c]
10
11
25
42
67[c]
100[c]
12
74
97
[a] GC yield. [b] With 5% of 1,3-diisopropylimidazolium chloride and
15% of KOtBu. [c] With 5% of 1,3-dicyclohexylimidazolium chloride
and 15% of KOtBu.
not contain an N-heterocyclic carbene and when applied di-
rectly in the amidation moderate yields of the product were
obtained (Table 3, entries 4 and 7). However, when 1,3-di-
isopropyl- or 1,3-dicyclohexylimidazol-2-ylidene were gener-
ated together with these complexes the yield of the amide
increased considerably (entries 5, 6, 8, and 9) and was com-
parable to the best results in Tables 1 and 2. The modified
Grubbs catalyst with the phenyl indenylidene ligand[29]
showed the same results (entries 10 and 11), which under-
lines the assumption that the benzylidene group in the meta-
thesis catalyst did not take part in the amidation, but was re-
duced off by the liberated dihydrogen. A number of other
N-heterocyclic carbenes were also generated together with
Grubbs 1st-generation catalyst,[30] but in all cases lower
yields of the amide was obtained. This confirms the results
in Table 1 that the imidazol-2-ylidine with 1,3-diisopropyl or
1,3-dicyclohexyl groups are the optimum N-heterocyclic car-
benes for the amidation. The in situ formation of the ruthe-
nium N-heterocyclic carbene complex was confirmed by pre-
paring the known cyclohexyl complex in Table 3, entry 12
from Grubbs 1st-generation catalyst.[31] When this well-de-
fined complex was applied in the amidation essentially the
same yield was obtained as when the complex was generat-
Chem. Eur. J. 2010, 16, 6820 – 6827
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6823

R. Madsen et al.
Table 4. Amidation of amines with primary alcohols.
water reservoir and 70 mL was
collected after 20 h. This corre-
sponds to 3 mmol and the gas
was shown to be dihydrogen by
GC analysis.
Entry
Alcohol
Amine
Amide
Yield [%][a] Yield [%][a] Yield [%][a]
cat. A
cat. B
cat. C
Mechanism: The amidation is
believed to proceed by forma-
tion of the aldehyde and the
hemiaminal as depicted in
Scheme 1. Esters are not inter-
mediates in the reaction, which
was confirmed by treating 2-
1
2
3
4
93[b]
95
92
100[b]
79
90
94
78
95
86
80
phenylethyl
with benzylamine and cata-
AHCTUNGTERGlNNUN yst A in refluxing toluene.
2-phenylacetate
78
Under these conditions, the
ester was stable and none of
the amide in Tables 1–3 was ob-
served. Imines are also highly
stable under the reaction condi-
tions, which was confirmed by
the reaction of N-benzylidene
benzylamine with catalyst A in
refluxing toluene. No conver-
sion of the imine occurred and
this did not change by adding
water or by conducting the re-
action under a dihydrogen at-
mosphere.
5
83[b]
71
73
6
7
8
60
90
70
82
93
85
78
87
78
9
60
65
53
68
44
60
10
Imines or reduction products
of imines have not been ob-
served as byproducts in any of
the experiments in Table 4 re-
gardless of the catalyst being
used. This may indicate that the
intermediate aldehyde stays co-
ordinated to the ruthenium cat-
alyst and is not released into
the reaction mixture. If this is
true, an externally added alde-
hyde may not be able to enter
11
49
53
48
12
13
21[c]
70[c]
35[c]
65[c]
33[c]
70[c]
[a] Isolated yield. [b] With 2% of [RuCl2ACTHNUTRGNEUNG(cod)], 2% of ligands and 8% of KOtBu. [c] In mesitylene at 1638C.
shows some sensitivity towards the steric demand around
the alcohol and the amine as well as additional coordinating
groups in the substrates. Attempts to use ammonia or am-
monia equivalents, such as LiNH2, NH4HCO3, Cu-
ACHTUNGTRENNUNG(NH3)4SO4·H2O, and MgCAHTUNTGREN(NGUN NH3)6Cl2, to afford a primary
amide failed completely and only gave the ester in various
amounts.
In most cases, the three different catalysts did not show
any major differences in yield and reactivity in Table 4. This
indicates that the catalytically active species is the same in
all three cases. For practical application, however, the most
convenient procedure is to generate the active catalyst in
situ. The evolution of dihydrogen was confirmed by repeat-
ing the experiment in entry 1 with 2 mmol of alcohol and
amine. The reaction flask was connected to a burette with a
the catalytic cycle and form the amide. To probe this ques-
tion a crossover experiment was carried out with p-methyl-
benzyl alcohol (1 equiv) and benzaldehyde (1 equiv), which
were reacted with n-hexylamine (2 equiv) in the presence of
complex 1 (5%) and potassium tert-butoxide (10%). Under
these conditions, the aldehyde was immediately converted
into the imine, whereas the alcohol reacted slowly to form
the corresponding imine (and not the amide) with about
50% conversion after 24 h. It appears that the imine from
the aldehyde inhibits formation of the amide from the alco-
hol causing the reaction to slow down and to stop at the
imine stage. A new experiment was therefore performed in
which benzaldehyde (1 equiv) was added over 3 h to a reac-
tion mixture with p-methylbenzyl alcohol (1 equiv), n-hexyl-
amine (2 equiv), complex 1 (5%), and potassium tert-butox-
6824
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6820 – 6827

Amide Synthesis from Alcohols and Amines
FULL PAPER
ide (10%). Although the amidation was slow in this case it
did not stop and the alcohol was converted into a 6:1 mix-
ture of the amide and the imine with almost complete con-
version after 30 h and with 50% conversion after 4 h. Again,
the aldehyde reacted immediately to produce the corre-
sponding imine, but in this case, a small amount of N-benzyl
benzamide was also observed as a byproduct. The ratio be-
tween the amide from the alcohol and the aldehyde was
10:1 after both 4 and 30 h. This does not indicate that a
crossover takes place to a significant degree and we, there-
fore, believe the intermediate aldehyde in the amidation
stays coordinated to the ruthenium catalyst.
In a previous study, ruthenium 1,3-diisopropylimidazol-2-
ylidene complex 4 was converted into the five-membered
ꢀ
ruthenacycle 5 by C H activation of the isopropyl methyl
group[33,34] (Scheme 3). The reaction was facilitated by hy-
Scheme 4. Proposed mechanism for amide formation.
plex 7. This part is similar to what has been established for
Scheme 3. Interconversion between 4 and 5[33] and the structure of 6.
ruthenium transfer hydrogenation catalysts.[37,38] It should,
however, be noted that [RuCl
with alcohols under basic conditions to form the dihydride
complex [RuH2A
(PPh3)3].[38] Whether complex 1 also reacts
2ACHTUNGTREN(NUNG PPh3)3] is known to react
drogen acceptors, such as olefins and could be reversed by
hydrogen donors, such as dihydrogen or alcohols.[33] It could
CHTUNGTRENNUNG
twice with the alkoxide is not known at this point. In fact,
the remaining ligand(s) on ruthenium in 7 could be chloride,
hydride, or an amine and is, therefore, denoted Ln in
Scheme 4. A more thorough mechanistic study will have to
be carried out to differentiate between these scenarios. With
formation of the aldehyde complex 7 a catalytic cycle can be
proposed for which the amine adds to the aldehyde to form
the hemiaminal, which stays coordinated to the metal. Re-
lease of hydrogen can take place by hydrogen transfer to hy-
dride as previously established.[39] This gives rise to complex
8, which upon b-hydride elimination releases the amide. Co-
ordination of the alcohol and a second hydrogen transfer to
hydride affords the alkoxide complex 9, which is ready to
re-enter the catalytic cycle. It should be noted that all the
ruthenium species in the catalytic cycle remain in the same
oxidation state as the starting complex. The added phos-
phine presumably stabilizes catalyst resting states and is not
believed to be involved in the catalytic cycle since the ami-
dation can be performed with a variety of phosphines and
other ligands.[19]
ꢀ
not be completely excluded that a similar C H activation
would take place with our 1,3-diisopropylimidazol-2-ylidene
ligand and thereby explain the high reactivity of this ligand
in the amidation. To probe this question experimentally, we
ꢀ
prepared deuterated complex 6. If C H activation of the
isopropyl methyl groups is a major reaction pathway we
would expect a significant amount of deuterium in the hy-
drogen gas from the reaction. However, when equimolar
amounts of 2-phenylethanol and benzylamine were treated
with 6 (10%), PCy3 (10%), and KHMDS (15%) in reflux-
ing toluene for 1 h, only a 51:1 ratio of H2 and H D was
measured by selected ion monitoring. This low amount of
H D does not indicate that a rapid exchange reaction is
taking place and, therefore, we do not believe that a C H
oxidative addition reaction is involved in the catalytic cycle.
The high reactivity of the 1,3-diisopropyl- and 1,3-dicyclo-
hexylimidazol-2-ylidene ligands is probably a result of the
relatively low steric demand of these ligands relative to
other N-heterocyclic carbene ligands.[35] Furthermore, in
ruthenium(II) complexes with these ligands agostic interac-
tions have been observed between ruthenium and the CH2/
CH3 hydrogen atoms on the ligand that may serve to further
stabilize coordinatively unsaturated species in the catalytic
cycle.[36]
ꢀ
ꢀ
ꢀ
Conclusion
In summary, we have presented an atom-economical proce-
dure for the direct synthesis of amides from alcohols and
amines in which dihydrogen is formed as the only byprod-
uct. The reaction is catalyzed by ruthenium N-heterocyclic
carbene complexes that are easy to handle and straightfor-
Based on these studies, we propose the reaction mecha-
nism in Scheme 4. The transformation is initiated by loss of
the p-cymene ligand upon heating. Reaction with an alkox-
ide followed by b-hydride elimination affords aldehyde com-
Chem. Eur. J. 2010, 16, 6820 – 6827
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6825

R. Madsen et al.
C25H38Cl2N2Ru: C 55.75, H 7.11, N 5.20; found: C 55.14, H 6.84, N 5.16;
1H NMR spectroscopic data are in accordance with literature values.[24c]
ward to modify. Three different catalyst systems have been
developed that show similar reactivity and yields in the ami-
dation with a wide variety of substrates. A mechanism is
proposed with ruthenium(II) N-heterocyclic carbene species
as the catalytically active components and for which the in-
termediate aldehyde and hemiaminal remain coordinated to
ruthenium in the catalytic cycle. The reaction presents a
new direction in the synthesis of one of the most important
linkages in organic chemistry.
General procedure for amidation with complex 1 (catalyst B): [RuCl2-
ACHUTNGERN(NUG IiPr)HCATUNGTREN(NNGU p-cymene)] (1) (11.5 mg, 0.025 mmol), PCy3 (7.0 mg, 0.025 mmol),
and KOtBu (5.6 mg, 0.05 mmol) were placed in an oven-dried Schlenk
tube. Vacuum was applied and the tube was then filled with argon (re-
peated twice). Freshly distilled toluene (1 mL) was added and the mix-
ture was heated to reflux under an argon atmosphere for 20 min. The al-
cohol (0.5 mmol) and the amine (0.5 mmol) were added and the mixture
was heated to reflux under an argon atmosphere for 24 h and then
worked up as described above.
General procedure for amidation with metathesis catalyst (catalyst C):
Hoveyda–Grubbs 1st-generation catalyst (15 mg, 0.025 mmol), 1,3-diiso-
propylimidazolium chloride (4.7 mg, 0.025 mmol), and KOtBu (8.4 mg,
0.075 mmol) were placed in an oven-dried Schlenk tube. Vacuum was ap-
plied and the tube was then filled with argon (repeated twice). Freshly
distilled toluene (1 mL) was added and the mixture was heated to reflux
under an argon atmosphere for 20 min. The alcohol (0.5 mmol) and the
amine (0.5 mmol) were added and the mixture was heated to reflux
under an argon atmosphere for 24 h and then worked up as described
above.
Experimental Section
General: Toluene was destilled from sodium and benzophenone under a
nitrogen atmosphere. NMR spectra were recorded on a Varian Mercury
300 Bruker AC 200 spectrometer while IR spectra were obtained on a
Bruker alpha-P spectrometer. Mass spectrometry was performed by
direct inlet on a Shimadzu-GCMS-QP5000 instrument of for hydrogen
analysis on a Pfeiffer OmniStar GSD 301. GC yields were obtained with
dodecane as internal standard on
a Shimadzu GC-2010 instrument
equipped with a Supelco Equity-1 capillary column (15 mmꢁ0.10 mm,
0.10 mm film). Microanalyses were obtained at the Microanalytical Labo-
ratory, University of Vienna.
Acknowledgements
General procedure for amidation with an in situ catalyst (catalyst A):
[21]
[RuCl2ACHTUNGTRENNUNG(cod)] (7.0 mg, 0.025 mmol), PCyp3·HBF4 (8.2 mg, 0.025 mmol),
We thank the Torkil Holm Foundation and the Danish National Re-
search Foundation for financial support.
1,3-diisopropylimidazolium chloride (4.7 mg, 0.025 mmol), and KOtBu
(11.2 mg, 0.10 mmol) were placed in an oven-dried Schlenk tube.
Vacuum was applied and the tube was then filled with argon (repeated
twice). Freshly distilled toluene (1 mL) was added and the mixture was
heated to reflux under an argon atmosphere for 20 min. The alcohol
(0.5 mmol) and the amine (0.5 mmol) were added and the mixture was
heated to reflux under an argon atmosphere for 24 h. The reaction mix-
ture was cooled to room temperature and the solvent removed in vacuo.
The residue was purified by silica-gel column chromatography (pentane/
EtOAc 4:1!1:1) to afford the amide.
[1] a) T. Cupido, J. Tulla-Puche, J. Spengler, F. Albericio, Curr. Opin.
Drug Discovery Dev. 2007, 10, 768–783; b) J. S. Carey, D. Laffan, C.
Thomson, M. T. Williams, Org. Biomol. Chem. 2006, 4, 2337–2347;
c) M. Okada, Prog. Polym. Sci. 2002, 27, 87–133.
[2] C. A. G. N. Montalbetti, V. Falque, Tetrahedron 2005, 61, 10827–
10852.
[3] a) R. M. Al-Zoubi, O. Marion, D. G. Hall, Angew. Chem. 2008, 120,
2918–2921; Angew. Chem. Int. Ed. 2008, 47, 2876–2879; b) T. Maki,
K. Ishihara, H. Yamamoto, Tetrahedron 2007, 63, 8645–8657.
[4] J. W. Comerford, J. H. Clark, D. J. Macquarrie, S. W. Breeden,
Chem. Commun. 2009, 2562–2564.
ACHTUNGTRENNUNG[RuCl2ACHTUNGTRENNUNG(IiPr)ACHTUNGTRENNUNG(p-cymene)] (1): 1,3-Diisopropylimidazolium chloride
(124.1 mg, 0.77 mmol) and Ag2O (75.3 mg, 0.33 mmol) were suspended in
anhydrous, degassed CH2Cl2 (7 mL) under argon and refluxed for 1 h in
a Schlenk flask with a reflux condenser. [RuCl2ACTHUNTRGNEUNG(p-cymene)]2 (201.0 mg,
0.33 mmol) in anhydrous, degassed CH2Cl2 (3 mL) was then added and
the solution was refluxed for 2 h and concentrated in vacuo. The residue
was purified on a short silica-gel column (CH2Cl2/iPrOH 9:1) to give
295.0 mg (98%) of a red/orange solid. Rf =0.64 (CH2Cl2/iPrOH 9:1); IR
(neat): n˜ =3152, 3099, 3077, 2958, 2930, 2870, 1473, 1412, 1391, 1369,
[5] E. Valeur, M. Bradley, Chem. Soc. Rev. 2009, 38, 606–631.
[6] a) D. Gnanamgari, R. H. Crabtree, Organometallics 2009, 28, 922–
924; b) N. A. Owston, A. J. Parker, J. M. J. Williams, Org. Lett. 2007,
9, 3599–3601.
[7] a) M. B. Soellner, B. L. Nilsson, R. T. Raines, J. Am. Chem. Soc.
2006, 128, 8820–8828; b) F. L. Lin, H. M. Hoyt, H. van Halbeek,
R. G. Bergman, C. R. Bertozzi, J. Am. Chem. Soc. 2005, 127, 2686–
2695.
[8] a) K. Ekoue-Kovi, C. Wolf, Chem. Eur. J. 2008, 14, 6302–6315;
b) W.-J. Yoo, C.-J. Li, J. Am. Chem. Soc. 2006, 128, 13064–13065.
[9] J. W. Bode, R. M. Fox, K. D. Baucom, Angew. Chem. 2006, 118,
1270–1274; Angew. Chem. Int. Ed. 2006, 45, 1248–1252.
[10] a) S. Lang, J. A. Murphy, Chem. Soc. Rev. 2006, 35, 146–156;
b) R. V. Kolakowski, N. Shangguan, R. R. Sauers, L. J. Williams, J.
Am. Chem. Soc. 2006, 128, 5695–5702.
1297, 1265, 1213, 1133, 856, 770, 700 cmꢀ1 1H NMR (300 MHz, CDCl3):
;
d=1.31 (d, J=6.9 Hz, 6H), 1.44 (brd, J=6.2 Hz, 12H), 2.08 (s, 3H), 2.92
(m, 1H), 5.15 (d, J=6.0 Hz, 2H), 5.31 (m, 2H), 5.47 (d, J=6.0 Hz, 2H),
7.07 ppm (s, 2H); 13C NMR (75 MHz, CDCl3): d=18.6, 22.8, 25.0, 30.8,
52.0, 83.4, 85.1, 97.1, 106.4, 118.9, 171.1 ppm; MS: m/z: calcd: 423.11
[MꢀCl]+; found: 423.07; elemental analysis calcd (%) for
C19H30Cl2N2Ru: C 49.78, H 6.60, N 6.11; found: C 49.84, H 6.44, N 6.05.
ACHTUNGTRENNUNG[RuCl2ACHTUNGTRENNUNG(ICy)ACHTUNGTRENNUNG(p-cymene)] (2): 1,3-Dicyclohexylimidazolium chloride
(200.2 mg, 0.75 mmol) and Ag2O (86.1 mg, 0.37 mmol) were suspended in
anhydrous, degassed CH2Cl2 (8 mL) under argon and refluxed for 1 h in
a Schlenk flask with a reflux condenser. [RuCl
(p-cymene)]2 (226.0 mg,
[11] W. Ye, J. Mo, T. Zhao, B. Xu, Chem. Commun. 2009, 3246–3248.
[12] a) J. W. Bode, S. S. Sohn, J. Am. Chem. Soc. 2007, 129, 13798–
13799; b) H. U. Vora, T. Rovis, J. Am. Chem. Soc. 2007, 129, 13796–
13797.
[13] a) A. Brennfꢂhrer, H. Neumann, M. Beller, Angew. Chem. 2009,
121, 4176–4196; Angew. Chem. Int. Ed. 2009, 48, 4114–4133; b) A.
Brennfꢂhrer, H. Neumann, M. Beller, ChemCatChem 2009, 1, 28–
41; c) Y. Uenoyama, T. Fukuyama, O. Nobuta, H. Matsubara, I.
Ryu, Angew. Chem. 2005, 117, 1099–1102; Angew. Chem. Int. Ed.
2005, 44, 1075–1078.
0.37 mmol) in anhydrous, degassed CH2Cl2 (3 mL) was then added and
the solution was refluxed for 1 h and concentrated in vacuo. The residue
was purified on a short silica-gel column (CH2Cl2/iPrOH 9:1) to give
368.4 mg (93%) of a red/orange solid. Rf =0.64 (CH2Cl2/iPrOH 9:1); IR
(neat): n˜ =3091, 2957, 2921, 2848, 1466, 1455, 1446, 1418, 1380, 1290,
1276, 1232, 1190, 897, 747, 697 cmꢀ1 1H NMR (300 MHz, CDCl3): d=
;
1.14–2.44 (m, 20H), 1.36 (d, J=6.9 Hz, 6H), 2.13 (s, 3H), 2.84 (m, 1H),
4.84 (m, 2H), 5.14 (d, J=6.0 Hz, 2H), 5.46 (d, J=6.0 Hz, 2H), 7.04 ppm
(s, 2H); 13C NMR (50 MHz, CDCl3): d=18.8, 23.1, 25.3, 25.4, 26.0, 31.2,
35.4, 35.8, 59.3, 83.6, 85.3, 97.3, 105.1, 119.3, 171.4 ppm; MS: m/z: calcd:
503.18 [MꢀCl]+; found: 503.15; elemental analysis calcd (%) for
[14] a) A. J. A. Watson, A. C. Maxwell, J. M. J. Williams, Org. Lett. 2009,
11, 2667–2670; b) T. Zweifel, J.-V. Naubron, H. Grꢂtzmacher,
6826
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 6820 – 6827

Amide Synthesis from Alcohols and Amines
FULL PAPER
Angew. Chem. 2009, 121, 567–571; Angew. Chem. Int. Ed. 2009, 48,
559–563.
[28] I. C. Stewart, T. Ung, A. A. Pletnev, J. M. Berlin, R. H. Grubbs, Y.
Schrodi, Org. Lett. 2007, 9, 1589–1592.
[15] For the synthesis of amides from primary alcohols and hydroxyla-
mine in the presence of styrene, see: N. A. Owston, A. J. Parker,
J. M. J. Williams, Org. Lett. 2007, 9, 73–75.
[16] C. Gunanathan, Y. Ben-David, D. Milstein, Science 2007, 317, 790–
792.
[29] A. Fꢂrstner, O. Guth, A. Dꢂffels, G. Seidel, M. Liebl, B. Gabor, R.
Mynott, Chem. Eur. J. 2001, 7, 4811–4820.
[30] The following 1,3-dialkylimidazolium salts were tested in the pres-
ence of Grubbs 1st-generation catalyst and base: alkyl=Me, tBu,
Mes, o-Tol, and p-Tol.
[17] L. U. Nordstrøm, H. Vogt, R. Madsen, J. Am. Chem. Soc. 2008, 130,
17672–17673.
[31] T. Weskamp, F. J. Kohl, W. A. Herrmann, J. Organomet. Chem.
1999, 582, 362–365.
[18] K.-i. Shimizu, K. Ohshima, A. Satsuma, Chem. Eur. J. 2009, 15,
9977–9980.
[32] The following primary amines were investigated: glycine tert-butyl
ester, glycinamide, and N-benzylglycinamide.
[19] See also: S. C. Ghosh, S. Muthaiah, Y. Zhang, X. Xu, S. H. Hong,
Adv. Synth. Catal. 2009, 351, 2643–2649.
[33] S. Burling, B. M. Paine, D. Nama, V. S. Brown, M. F. Mahon, T. J.
Prior, P. S. Pregosin, M. K. Whittlesey, J. M. J. Williams, J. Am.
Chem. Soc. 2007, 129, 1987–1995.
[20] The following imidazolium salts gave low yields in the amidation
with PCy3: 1,3-bis(2,6-diisopropylphenyl)imidazolium chloride, 1,3-
diisopropyl-4,5-dimethylimidazolium chloride, 1,3-diisopropylbenzi-
midazolium bromide, and 1,3-dibenzylimidazolium bromide.
[21] J. Zhou, G. C. Fu, J. Am. Chem. Soc. 2003, 125, 12527–12530.
[22] V. Dragutan, I. Dragutan, L. Delaude, A. Demonceau, Coord.
Chem. Rev. 2007, 251, 765–794.
[34] For a similar observation with a 1,3-dicyclohexylimidazol-2-ylidene
ruthenium complex, see: A. E. W. Ledger, M. F. Mahon, M. K.
Whittlesey, J. M. J. Williams, Dalton Trans. 2009, 6941–6947.
[35] R. Dorta, E. D. Stevens, N. M. Scott, C. Costabile, L. Cavallo, C. D.
Hoff, S. P. Nolan, J. Am. Chem. Soc. 2005, 127, 2485–2495.
[36] a) S. Burling, E. Mas-Marzꢄ, J. E. V. Valpuesta, M. F. Mahon, M. K.
Whittlesey, Organometallics 2009, 28, 6676–6686; b) S. Burling, G.
Kociok-Kçhn, M. F. Mahon, M. K. Whittlesey, J. M. J. Williams, Or-
ganometallics 2005, 24, 5868–5878.
[37] a) E. Solari, S. Gauthier, R. Scopelliti, K. Severin, Organometallics
2009, 28, 4519–4526; b) B. Martꢅn-Matute, J. B. ꢆberg, M. Edin, J.-
E. Bꢇckvall, Chem. Eur. J. 2007, 13, 6063–6072.
[38] A. Aranyos, G. Csjernyik, K. J. Szabꢃ, J.-E. Bꢇckvall, Chem.
Commun. 1999, 351–352.
[23] L. Delaude, S. Delfosse, A. Richel, A. Demonceau, A. F. Noels,
Chem. Commun. 2003, 1526–1527.
[24] a) L. Jafarpour, J. Huang, E. D. Stevens, S. P. Nolan, Organometallics
1999, 18, 3760–3763; b) H. Kꢂcꢂkbay, B. Cetinkaya, S. Guesmi,
P. H. Dixneuf, Organometallics 1996, 15, 2434–2439; c) W. A. Herr-
mann, C. Kçcher, L. J. Gooßen, G. R. J. Artus, Chem. Eur. J. 1996,
2, 1627–1636.
[25] a) P. Csabai, F. Joꢃ, Organometallics 2004, 23, 5640–5643;
b) H. M. J. Wang, I. J. B. Lin, Organometallics 1998, 17, 972–975.
[26] J. Louie, C. W. Bielawski, R. H. Grubbs, J. Am. Chem. Soc. 2001,
123, 11312–11313.
[39] S. L. Chatwin, M. G. Davidson, C. Doherty, S. M. Donald, R. F. R.
Jazzar, S. A. Macgregor, G. J. McIntyre, M. F. Mahon, M. K. Whittle-
sey, Organometallics 2006, 25, 99–110.
[27] For a recent review on nonmetathesis reactions with Grubbs ruthe-
nium carbenes, see: B. Alcaide, P. Almendros, A. Luna, Chem. Rev.
2009, 109, 3817–3858.
Received: March 4, 2010
Published online: April 30, 2010
Chem. Eur. J. 2010, 16, 6820 – 6827
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
6827