Amidation of Carboxylic Acids with Amines by Nb2O5 as a Reusable Lewis Acid Catalyst

Article abstract of DOI:10.1002/cctc.201500672

Among 28 types of heterogeneous and homogenous catalysts tested, Nb₂O₅ shows the highest yield for direct amidation of n-dodecanoic acid with a less reactive amine (aniline). The catalytic amidation by Nb₂O₅ is applicable to a wide range of carboxylic acids and amines with various functional groups, and the catalyst is reusable. A comparison of the results of the catalytic study and an infrared study of the acetic acid adsorbed on the catalyst suggests that activation of the carbonyl group of the carboxylic acid by Lewis acid sites on Nb₂O₅ is responsible for the high activity of the Nb₂O₅ catalyst. Kinetic studies show that Lewis acid sites on Nb₂O₅ are more water-tolerant than conventional Lewis acidic oxides (Al₂O₃, TiO₂). In comparison with the state-of-the-art homogeneous Lewis acid catalyst for amidation (ZrCl₄), Nb₂O₅ undergoes fewer negative effects from basic additives in the solution, which indicates that Nb₂O₅ is a more base-tolerant Lewis acid catalyst than the homogeneous Lewis acid catalyst.

Full text of DOI:10.1002/cctc.201500672

DOI: 10.1002/cctc.201500672
Full Papers
Amidation of Carboxylic Acids with Amines by Nb2O5 as
a Reusable Lewis Acid Catalyst
Md. A. Ali,[a] S. M. A. H. Siddiki,[b] Wataru Onodera,[a] Kenichi Kon,[a] and Ken-ichi Shimizu*[a, b]
Among 28 types of heterogeneous and homogenous catalysts
tested, Nb2O5 shows the highest yield for direct amidation of
n-dodecanoic acid with a less reactive amine (aniline). The cat-
alytic amidation by Nb2O5 is applicable to a wide range of car-
boxylic acids and amines with various functional groups, and
the catalyst is reusable. A comparison of the results of the cat-
alytic study and an infrared study of the acetic acid adsorbed
on the catalyst suggests that activation of the carbonyl group
of the carboxylic acid by Lewis acid sites on Nb2O5 is responsi-
ble for the high activity of the Nb2O5 catalyst. Kinetic studies
show that Lewis acid sites on Nb2O5 are more water-tolerant
than conventional Lewis acidic oxides (Al2O3, TiO2). In compari-
son with the state-of-the-art homogeneous Lewis acid catalyst
for amidation (ZrCl4), Nb2O5 undergoes fewer negative effects
from basic additives in the solution, which indicates that Nb2O5
is a more base-tolerant Lewis acid catalyst than the homoge-
neous Lewis acid catalyst.
Introduction
Amide bonds constitute the building blocks of pharmaceutical-
ly and biologically important compounds.[1–4] Conventionally,
amides are prepared from carboxylic acids and amines via acti-
vated carboxylic acid derivatives such as carboxylic acid anhy-
drides or acyl chlorides or through activation with a stoichio-
metric amount of a condensation agent for activation of the
carboxylic acid and water removal.[4] The conventional meth-
ods suffer from low atom efficiency and the production of by-
products. It is generally accepted that the catalytic synthesis of
amides from readily available starting materials is a priority
area for the pharmaceutical industry.[1] As summarized in
recent review articles,[5–9] the direct condensation of carboxylic
acids and amines by boron-based[10–19] or metal-based[20–26] ho-
mogeneous catalysts and oxide-based heterogeneous cata-
lysts[27–40] plays a central role in direct amidation methods.
However, less reactive amines such as anilines and less reactive
carboxylic acids such as a-hydroxycarboxylic acids and benzoic
acids are not generally tolerated by the previous catalysts. A
rare example is a boronic acid catalyst developed by Ishihara
and co-workers,[13] who have shown that condensation of an
equimolar mixture of a-hydroxycarboxylic acids and primary or
secondary amines proceeds with a boron-based catalyst under
azeotropic reflux conditions in toluene. However, homogene-
ous catalytic methods have the drawback of difficulties in cata-
lyst/product separation and catalyst reuse. Moreover, Lewis
acidic homogenous catalysts have potential drawbacks such as
the suppression of activity by strong coordination of basic
functional groups in a substrate (such as heterocyclic groups)
and the irreversible decomposition of the catalyst by water (as
a byproduct). As for heterogeneous catalysts for direct amida-
tion, previous reports mainly studied the N-formylation[4,32,35]
or N-acetylation[4,39] of amines. Some of the previous heteroge-
neous systems for amidation suffer from the drawbacks of lim-
ited scope and the need for excess amounts of reagent or
a special reaction method (microwave heating).[33,34]
In the course of our continuous studies on amide-bond-
forming reactions by heterogeneous Lewis acidic catalysts,[41–43]
we have recently reported that Nb2O5, prepared by calcination
of a commercial niobic acid, acts as a base-tolerant Lewis acid
catalyst for the direct imidation of dicarboxylic acids with
amines[43] and the direct amidation of esters with amines.[42]
We report herein that Nb2O5 is an effective and reusable cata-
lyst for the direct condensation of less reactive carboxylic acids
with less reactive amines. The catalytic results show wide ap-
plicability of the synthetic method, and IR spectroscopic and
kinetic studies show that the high activity of Nb2O5 is a result
of activation of the carboxylic acids by the Lewis acid sites of
Nb2O5 with a base-tolerant nature.
[a] M. A. Ali, W. Onodera, Dr. K. Kon, Prof. K.-i. Shimizu
Catalysis Research Center
Hokkaido University
N-21, W-10, Sapporo 001-0021 (Japan)
E-mail: kshimizu@cat.hokudai.ac.jp
Experimental Section
General
[b] Dr. S. M. A. H. Siddiki, Prof. K.-i. Shimizu
Elements Strategy Initiative for Catalysts and Batteries
Kyoto University
Commercially available organic compounds (from Tokyo Chemical
Industry or Aldrich) were used without further purification. GC (Shi-
madzu GC-2014) and GC–MS (Shimadzu GCMS-QP2010) analyses
were performed with an Ultra ALLOY+-1 capillary column (Frontier
Laboratories Ltd.) and with N2 and He as the carrier. All reactions
Katsura, Kyoto 615-8520 (Japan)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201500672.
ChemCatChem 2015, 7, 3555 – 3561
3555
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
were performed in oven-dried glassware under an inert atmos-
phere of nitrogen. Analytical TLC was performed on a Merck 60
F254 silica gel (0.25 mm thickness). Column chromatography was
performed with silica gel 60 (spherical, 63–210 mm, Kanto Chemical
Co. Ltd.). Molecular sieves (4 Š) was dehydrated at T=1008C.
then added to Nb2O5, followed by introduction of pyridine
(0.3 mmolgÀ1), purging with He for 600 s, and IR measurement of
the adsorbed species at T=2008C.
Catalytic tests
We used the as-received solvent without dehydration. The hetero-
geneous catalysts, stored under ambient conditions, were used for
catalytic reactions without any pretreatment, and thus, the catalyst
surface was hydrated before the reaction.
Catalyst preparation
Niobic acid (Nb2O5·nH2O, HY-340) was kindly supplied by CBMM.
Nb2O5 (surface area=54 m2 gÀ1) was prepared by calcination of
niobic acid at T=5008C for 3 h. MgO (JRC-MGO-3), TiO2 (JRC-TIO-
4), CeO2 (JRC-CEO-3), H+-type Y zeolite (HY) with a SiO2/Al2O3 ratio
of 4.8 (JRC-Z-HY-4.8), H+-type BEA zeolite (HBEA) with a SiO2/Al2O3
ratio of 25 (JRC-Z-HB25), and H+-type MFI zeolite (HMFI) with
a SiO2/Al2O3 ratio of 90 (JRC-Z5-90H) were supplied from the Catal-
ysis Society of Japan. SiO2 (Q-10, 300 m2 gÀ1) was supplied from
Fuji Silysia Chemical Ltd. ZrO2·nH2O was prepared by hydrolysis of
zirconium oxynitrate 2-hydrate in water with aqueous NH4OH solu-
tion, followed by filtration of the precipitate, washing with water
three times, and drying at T=2008C. ZrO2, ZnO, SnO2, MoO3, WO3,
Ta2O5, and CaO were prepared by calcination (T=5008C, t=3 h) of
the hydrous oxides ZrO2·nH2O, ZnO·nH2O (Kishida Chemical),
H2SnO3 (Kojundo Chemical Laboratory Co., Ltd.), H2MoO4 (Kanto
Chemical), H2WO4 (Kanto Chemical), Ca(OH)2 (Kanto Chemical), and
Ta2O5·nH2O (Mitsuwa Chemicals). g-Al2O3 and q-Al2O3 were pre-
pared by calcination of g-AlOOH (Catapal B Alumina purchased
from Sasol) for 3 h at T=900 and 10008C, respectively. Montmoril-
lonite K10 clay and sulfonic resins (Amberlyst-15 and Nafion-SiO2
composite) were purchased from Aldrich. Fe3+-exchanged K10 clay
(Fe3+-mont) was prepared by treating the clay with an aqueous so-
lution of FeCl3·6H2O for 3 h at RT, followed by centrifugation, wash-
ing with deionized water four times, and drying in vacuo at RT. The
Fe content in Fe3+-mont (0.46 wt%) was determined by inductively
coupled plasma atomic emission spectroscopy (ICP-AES) analysis.
Scandium(III) trifluoromethanesulfonate, Sc(OTf)3, was purchased
from Tokyo Chemical Industry. ZrCl4 was purchased from WAKO.
Cs2.5H0.5PW12O40 was prepared by titrating H3PW12O40 (Nippon Inor-
ganic Color and Chemicals Co.) with an aqueous solution of Cs2CO3
(0.10 moldmÀ3) with vigorous stirring, followed by centrifuging
and drying at T=2008C.
Typically, carboxylic acid (1 mmol) and amine (1 mmol) in toluene
(2 mL) and Nb2O5 (50 mg) were added to a reaction vessel (pyrex
cylinder) with a reflux condenser and a magnetic stirrer. The reac-
tion mixture was heated to reflux under an N2 atmosphere and
stirred at 400 rpm. For azeotropic removal of water, a funnel con-
taining 4 Š molecular sieves (0.2 g) on a cotton plug was placed in
the top of the cylinder surmounted by a reflux condenser. After
completion of the reaction, 2-propanol (4 mL) was added to the
mixture, and the Nb2O5 catalyst was separated by centrifugation.
For the catalytic tests in Table 1 and Figures 1, 4, 5, and 6, the reac-
tion mixture was analyzed by GC and the yield of the products
was determined by using n-dodecane as an internal standard. For
the reactions in Tables 3–5, the product was isolated by column
chromatography and the resulting product was then identified by
1
using GC–MS and H NMR and 13C NMR spectroscopy analyses.
Results and Discussion
Catalyst screening
We performed a model reaction between an equimolar
amount of n-dodecanoic acid and aniline under azeotropic
reflux conditions. Table 1 summarizes the yield of the corre-
sponding amide with various catalysts including metal oxides
and standard heterogeneous and homogeneous acid catalysts.
Figure 1 shows the time–yield profiles for some representative
catalysts. It is known that the direct formation of amides from
reactive amines and carboxylic acids without catalyst occurs in
nonpolar solvents under azeotropic reflux conditions.[21,43,44]
For the model reaction reported in Table 1, we used aniline, as
one of the least reactive amines in the literature, for the ther-
mal amidation reaction.[21,44] We confirmed that the thermal re-
action in the absence of catalyst gave only 1% yield of the
amide (Table 1, entry 1). We screened 17 types of simple metal
oxides (Table 1, entries 2–19) including two of the hydrates
(Table 1, entries 4 and 11). Among the oxides tested, Nb2O5
showed the highest yield (99%) of the amide. In the literature,
TiO2,[33] ZnO,[32] Al2O3,[35,36] ZrO2·nH2O,[38] SiO2,[27] Fe3+-mont,[39,40]
and the HY[28] and HBEA[29] zeolites were reported to be effec-
tive for the direct amidation. However, these catalysts showed
lower yields than Nb2O5. For example, conventional solid Lewis
acids[45,46] such as TiO2, alumina, and Fe3+-mont (Table 1, en-
tries 5, 8, 9, and 20) gave low to moderate yields (9–66%).
Basic oxides (MgO, CaO) were ineffective. In the dehydrative
amide condensation reaction, water produced during the reac-
tion can suppress the catalytic activity by strong adsorption on
acid sites of catalysts. Thus, water-tolerant acid catalysts may
be effective for the reaction. We tested water-tolerant Brønsted
acidic heterogeneous catalysts,[47] such as a high-silica zeolite
In situ IR
In situ IR spectra were recorded by a JASCO FT/IR-4200 spectrome-
ter equipped with a mercury cadmium telluride detector. For the
acetic acid adsorption IR study, a closed IR cell surrounded by
a Dewar vessel was connected to an evacuation system. During
the IR measurement, the IR cell was cooled by a freezing mixture
of ethanol/liquid nitrogen in the Dewar vessel, and the thermocou-
ple near the sample showed T=(À75Æ5)8C. The sample was
pressed into
a self-supporting wafer (40 mg, 1=2 cm) and
mounted into the IR cell with CaF2 windows. Spectra were mea-
sured by accumulating 15 scans at a resolution of 4 cmÀ1. After in
situ pre-evacuation of the sample at T=5008C for 0.5 h, a reference
spectrum of the sample disc was measured at T=(À75Æ5)8C. The
sample was then exposed to acetic acid (200 Pa) at T=(À75Æ5)8C
for 120 s, followed by evacuation for 500 s. A differential IR spec-
trum, with respect to the reference spectrum, was then recorded
at T=(À75Æ5)8C. The pyridine-adsorption IR study was performed
at T=2008C with a flow-type IR cell connected to a flow reaction
system. The IR disc of Nb2O5 in the IR cell was first dehydrated
under an He flow at T=5008C, and then a background spectrum
was taken under an He flow at T=2008C. H2O (1.4 mmolgÀ1) was
ChemCatChem 2015, 7, 3555 – 3561
www.chemcatchem.org
3556
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
acid (Table 1, entry 4), which has been studied as a water-toler-
ant Lewis acid catalyst,[49] gave a lower yield (74%) than Nb2O5.
Table 1. Catalyst screening for amidation of n-dodecanoic acid with
aniline.
Lewis acidity of Nb2O5
In our previous IR study of pyridine adsorption on Nb2O5, we
showed that the surface acid sites of dehydrated Nb2O5 are
mainly Lewis acidic sites (exposed Nb5+ cations).[45] Figure 2
Entry
Catalyst
Yield [%][a]
1
2
3[b]
4
5
6
7
8
no catalyst
Nb2O5
Nb2O5
niobic acid
TiO2
ZnO
1
99
96
74
51
21
16
11
9
WO3
q-Al2O3
g-Al2O3
MoO3
9
10
11
12
13
14
15
16
17
19
20
21
22
23
24
25
26
27
28
29[c]
30
9
ZrO2·nH2O
CeO2
ZrO2
Ta2O5
CaO
SnO2
MgO
SiO2
Fe3+-mont
HY
9
7
6
6
5
5
1
13
66
7
27
26
7
4
30
2
Figure 2. IR spectra of pyridine adsorbed on dehydrated Nb2O5 (dashed line)
and rehydrated Nb2O5 (solid line) at T=2008C.
HBEA
HMFI
Cs2.5H0.5PW12O40
Amberlyst-15
Nafion-SiO2
Sc(OTf)3
ZrCl4
shows the IR spectrum of pyridine adsorbed on dehydrated
and rehydrated Nb2O5. These spectra have basically the same
features; the band at n˜ =1445 cmÀ1 as a result of coordinated
pyridine on Lewis acid sites (exposed Nb5+ cations) is domi-
nant rather than the band at n˜ =1540 cmÀ1 as a result of pyri-
dinium ions because of Brønsted acid sites. The result shows
that water does not essentially change the IR spectrum of ad-
sorbed pyridine; Nb2O5 is predominantly Lewis acidic even
after rehydration. To investigate the Lewis acid–base interac-
tion between the Nb site and a carbonyl group of a model car-
boxylic acid, we measured the in situ IR spectrum of acetic
acid adsorbed on Nb2O5. The spectrum (Figure 3) showed a
71
1
19
H2SO4
PTSA[d]
[a] Yields measured by GC analysis. [b] Under reflux conditions without
azeotropic water removal for t=40 h. [c] Aqueous solution of H2SO4
(30 wt%). [d] PTSA: p-Toluenesulfonic acid.
C=O stretching band of the adsorbed acetic acid (nC O) at
=
Figure 1. Time–yield profiles for amidation of n-dodecanoic acid (1 mmol)
with aniline (1 mmol) catalyzed by metal oxides (50 mg) under different con-
ditions: *: azeotropic reflux; *: reflux (without azeotropic water removal);
!: reflux with 3 mmol of H2O in the initial mixture.
(HMFI), Cs-exchanged heteropoly acid, and the acidic resins
Amberlyst-15 and Nafion-SiO2 (Table 1, entries 23–26, respec-
tively), and also the water-tolerant homogeneous Lewis acid[48]
Sc(OTf)3 (Table 1, entry 27). However, these water-tolerant acid
catalysts gave lower yields of the amide (2—30%) than Nb2O5.
Homogenous Brønsted acids such as sulfuric acid and p-
toluenesulfonic acid (PTSA; Table 1, entries 29 and 30, respec-
tively) also gave low yields. A hydrate of Nb2O5 called niobic
Figure 3. IR spectra of acetic acid adsorbed on Nb2O5, TiO2, q-Al2O3, and SiO2
at T=À758C.
ChemCatChem 2015, 7, 3555 – 3561
www.chemcatchem.org
3557
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
a lower wavenumber (n˜ =1686 cmÀ1) than the non Lewis
acidic oxide SiO2 (n˜ =1703 cmÀ1) and conventional Lewis acidic
oxides (TiO2: n˜ =1695 cmÀ1; Al2O3: n˜ =1697 cmÀ1). This indi-
cates that the surface of Nb2O5 has the most effective Lewis
acid sites for activation of the C=O bond of the carboxylic acid.
Base-tolerant catalysis by Nb2O5
Lewis acidic catalysts for direct amidation should work even in
the presence of water, because the reaction yields water as
a coproduct. We studied the effect of water removal and water
addition on the time–yield profiles for some Lewis acidic metal
oxide catalysts (Nb2O5, TiO2, and Al2O3) for the model amida-
tion of n-dodecanoic acid and aniline (Figure 1). For all of the
catalysts, the standard azeotropic reflux conditions gave higher
activity than the reaction without azeotropic water removal,
and the reaction without azeotropic water removal with
3 mmol of H2O in the initial mixture gave the lowest activity.
However, the negative impact of the water was lower for
Nb2O5 than for TiO2 and Al2O3. As shown in Figure 4, the initial
Figure 5. Yield of amide for the reaction of n-dodecanoic acid with aniline
for 30 h with Nb2O5 (50 mg, 0.38 mmol) and ZrCl4 (50 mg, 0.38 mmol) in the
absence and presence of the basic additives (0.5 mmol): water, 2,6-dimethyl-
pyridine, pyridine, and triethylamine.
tions gave higher yields for both catalysts, but the negative ef-
fects of the additives were lower for Nb2O5 than for ZrCl4. No-
tably, the 0.5 mmol of basic molecules added to the mixture is
172 times larger than the number of surface Lewis acid sites
on the Nb2O5 catalyst used. This suggests that the active site
(the Nb5+ Lewis acid site) interacts preferentially with the reac-
tant (carboxylic acid) in the presence of an excess amount of
basic molecules. To summarize the above results, we can con-
clude that the Lewis acid site of Nb2O5 has higher tolerance to
basic molecules than conventional solid Lewis acids and a typi-
cal homogeneous Lewis acid. The water-tolerant character of
the Nb5+ Lewis acid sites of Nb2O5 is consistent with the pio-
neering work by Nakajima et al.[49]
Performance of Nb2O5-catalyzed amidation
Figure 4. Initial rate for amidation of n-dodecanoic acid with aniline by
Nb2O5 (*) and TiO2 (*) as a function of the initial concentration of water
As listed in Table 2, the turnover number (TON) with respect to
the Lewis acid site of Nb2O5 (341) was more than 200 times
higher than those of ZrCl4 (a well-established homogenous cat-
alyst for direct amidation[20–22]) and Sc(OTf)3 (a well-established
“water-tolerant” Lewis acid[48]). The TON of Nb2O5 was five
times larger than that of TiO2. As discussed in the above sec-
tion, the higher catalytic efficiency of Nb2O5 can be attributed
(CH O =1.1 to 3.4m).
2
rate of amide formation with Nb2O5 and TiO2 decreased with
an increase in the initial concentration of water. This indicates
that water inhibits the activity of these catalysts. The slope was
lower for Nb2O5 than TiO2, and the reaction orders with respect
to water were À0.3 and À1.8 for Nb2O5 and TiO2, respectively.
This indicates that the water tolerance of Nb2O5 is higher than
that of TiO2.
Table 2. Summary of IR and kinetic results.
[b]
[c]
Catalyst
LA[a]
[mmolgÀ1
nC
nH O
TOF[d]
[hÀ1
TON[d]
=
O
2
]
[cmÀ1
]
]
ZrCl4 is a well-established Lewis acidic homogenous catalyst
for direct amidation.[20–22] Generally, the activity of homogene-
ous Lewis acids can be reduced by water and organic bases.
To compare the base tolerances of ZrCl4 and Nb2O5, we mea-
sured the yield of the amide in the standard reaction for t=
30 h with ZrCl4 or Nb2O5 under the azeotropic reflux conditions
in the absence or presence of basic additives (0.5 equiv.): H2O,
2,6-dimethylpyridine, pyridine, and triethylamine (Figure 5). Al-
though we used the same molar amount of the catalyst
(0.38 mmol), ZrCl4 was dissolved in the reaction mixture,
whereas Nb2O5 was insoluble. Clearly, the additive-free condi-
Nb2O5
TiO2
ZrCl4
0.058
0.083
4.3
1686
1695
–
À0.4
À1.8
–
11.4
2.0
0.006
0.003
341
61
1.7
0.1
Sc(OTf)3
2.0
–
–
[a] The number of Lewis acid sites on the surface of oxides Nb2O5 and
TiO2 was estimated by pyridine adsorption at T=2008C, as reported in
ref. [45]. The values for ZrCl4 and Sc(OTf)3 are based on the molecular
weights of the salts. [b] Position of the nC IR band of adsorbed acetic
=
O
acid (Figure 4). [c] Reaction order with respect to water (Figure 2).
[d] TOF: Turnover frequency; TON: turnover number. Calculated from the
number of Lewis acid sites and the catalytic results in Table 1.
ChemCatChem 2015, 7, 3555 – 3561
www.chemcatchem.org
3558
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
to the higher water tolerance and more effective Lewis acid ac-
tivation of the C=O bond by Nb2O5 than TiO2. Notably, the
water-tolerance of Nb2O5 enabled the amidation without azeo-
tropic water removal; the reaction with Nb2O5 under simple
reflux conditions for 40 h resulted in a 96% yield of the amide
(Table 1, entry 3).
Table 3. Nb2O5-catalyzed amidation of n-dodecanoic acid with various
amines.
We studied the reusability of Nb2O5. After the standard reac-
tion (Table 1, entry 2), the catalyst was separated from the mix-
ture by centrifugation, washed with acetone, and dried at T=
908C for 3 h. ICP-AES analysis of the solution confirmed that
the content of Nb in the solution was below the detection
limit. The recovered catalyst was reused five times without
a marked loss of its catalytic activity (Figure 6). For the stan-
Entry
1
Amine
Product
Yield[a] [%]
98
2
98
97
98
80
97
3
4
5[b]
6[b]
7
8
9
88
80
94
Figure 6. Reusability of Nb2O5 (50 mg) for amidation of n-dodecanoic acid
(1 mmol) with aniline (1 mmol) in toluene heated to reflux for 30 h.
dard reaction, the reaction was completely terminated by re-
moving the Nb2O5 catalyst from the reaction mixture after 4 h
(at 19% yield), and further heating of the filtrate for 26 h did
not increase the yield. These results indicate that Nb2O5 acts as
a reusable heterogeneous catalyst.
We then explored the generality and scope of the Nb2O5-cat-
alyzed direct amidation of carboxylic acids with different
amines (Tables 3–5). As listed in Table 3, anilines (entries 1–6)
with electron-donating and electron-withdrawing functional
groups, benzyl amines (entries 7–9) with electron-rich and elec-
tron-poor rings, heteroaromatic amines (entry 10), and aliphatic
primary amines (entries 11–13) with various functional groups
(phenyl, C=C, and hydroxy groups) all reacted with an equimo-
lar amount of n-dodecanoic acid to give the corresponding
amide in good to high yields (80–98%) after isolation. As
a result of the low nucleophilicity, the least reactive amines, Br-
and Cl-substituted anilines (Table 3, entries 5 and 6) and allyl-
amine (Table 3, entry 12) required higher temperature (heating
to reflux in o-xylene). A secondary amine, morpholine (Table 3,
entry 14) was also tolerated to give the corresponding tertiary
amide in high yield.
10
11
95
83
12[b]
13
98
96
14
81
[a] Yields of isolated products. [b] In o-xylene heated to reflux.
Table 4 shows that the method is also effective for the ami-
dation of various carboxylic acids with the less nucleophilic
amine aniline. Linear aliphatic carboxylic acids (Table 4, en-
tries 1–5) and the less reactive benzoic acid (Table 4, entry 6)
were converted into the corresponding amides in good to
high yields (81–98%) after isolation. The amidation of a hetero-
aromatic carboxylic acid, pyridine-2-carboxylic acid (Table 4,
entry 7), with benzylamine was also successful to give 90%
yield of the product.
Finally, we tested the amidation of more challenging carbox-
ylic acids (a-hydroxy-, b-hydroxy-, and b-thiocarboxylic acids)
with various amines under azeotropic reflux in o-xylene
(Table 5). Notably, there is only one report by Ishihara and co-
workers for the successful direct amidation of a-hydroxycar-
ChemCatChem 2015, 7, 3555 – 3561
www.chemcatchem.org
3559
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
boxylic acids with amines, but the previous method with
MeB(OH)2 as the catalyst is not effective for less reactive
amines such as aniline.[13] To our delight, our method was ap-
plicable to the synthesis of amides from aniline and a-hydroxy-
carboxylic acids (Table 5, entries 1 and 3) and a b-thiocarboxylic
acid (Table 5, entry 6). The a-hydroxycarboxylic acids include
an important biomass-derived chemical, lactic acid (Table 5, en-
tries 1 and 2), which demonstrates that our method can con-
tribute to production of fine chemicals from biomass feed-
stock. The method was also effective for the amidation of a b-
hydroxycarboxylic acid, salicylic acid (Table 5, entry 5), with
benzylamine and gave the corresponding amide in 95% yield.
We tentatively assume that the unprecedentedly efficient catal-
ysis of Nb2O5 for the amidation of challenging substrates is
caused by the base-tolerant Lewis acid activation of carboxylic
acids, which is evidenced by IR spectroscopy (Figures 2 and 3)
and kinetic studies (Figures 1, 4, and 6).
Table 4. Nb2O5-catalyzed amidation of various carboxylic acids with
aniline or benzylamine.
Entry
1
Acid
Product
Yield[a] [%]
80
2
83
81
98
95
3
4[b]
Conclusions
5
We have presented a versatile and sustainable method for the
direct amidation of carboxylic acids with various amines by
using Nb2O5 as a reusable, inexpensive, and commercially avail-
able heterogeneous catalyst. This simple and atom-efficient
method tolerates various functional groups and is applicable
to challenging substrates such as anilines and a-hydroxycar-
boxylic acids. The Lewis acid site of Nb2O5, as the active site
for the amidation, has higher tolerance to the copresent basic
molecules (water and tertiary and heteroaromatic amines) than
the state-of-the-art homogeneous Lewis acid catalyst for the
amidation (ZrCl4) and conventional Lewis acidic heterogeneous
catalysts (Al2O3 and TiO2), which results in higher catalytic activ-
ity of Nb2O5 than of these catalysts.
6[b]
80
90
7[b]
[a] Yields of isolated products. [b] In o-xylene heated to reflux.
Table 5. Nb2O5-catalyzed amidation of a-hydroxy-, b-hydroxy-, and
b-thiocarboxylic acids with various amines.[a]
Acknowledgements
Entry
1
Acid
Product
Yield[b] [%]
This work was supported by Grant-in-Aids for Scientific Re-
search B (26289299) from MEXT (Japan), a MEXT program “Ele-
ments Strategy Initiative to Form Core Research Center”, and
a Grant-in-Aid for Scientific Research on Innovative Areas “Nano
Informatics” (25106010) from JSPS.
69
2
3
4
80
71
65
Keywords: amines
·
carboxylic acids
·
heterogeneous
catalysis · Lewis acids · niobium
[1] D. J. C. Constable, P. J. Dunn, J. D. Hayler, G. R. Humphrey, J. J. L. Leazer,
R. J. Linderman, K. Lorenz, J. Manley, B. A. Pearlman, A. Wells, A. Zaks,
T. Y. Zhang, Green Chem. 2007, 9, 411 –420.
[2] S. D. Roughley, A. M. Jordan, J. Med. Chem. 2011, 54, 3451–3479.
[3] J. S. Carey, D. Laffan, C. Thomson, M. T. Williams, Org. Biomol. Chem.
2006, 4, 2337–2347.
[4] E. Valeur, M. Bradley, Chem. Soc. Rev. 2009, 38, 606–631.
[5] H. Lundberg, F. Tinnis, N. Selander, H. Adolfsson, Chem. Soc. Rev. 2014,
43, 2714–2742.
5
6
95
87
´
[6] E. Dimitrijevic, M. S. Taylor, ACS Catal. 2013, 3, 945–962.
[7] R. M. Lanigan, T. D. Sheppard, Eur. J. Org. Chem. 2013, 7453–7465.
[8] H. Charville, D. Jackson, G. Hodges, A. Whiting, Chem. Commun. 2010,
46, 1813–1823.
[a] Reaction conditions: Carboxylic acid (1 mmol), amine (1 mmol),
o-xylene (2 mL), azeotropic reflux, t=30 h. [b] Yields of isolated products.
[9] K. Ishihara, Tetrahedron 2009, 65, 1085–1109.
ChemCatChem 2015, 7, 3555 – 3561
www.chemcatchem.org
3560
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers
[10] K. Ishihara, S. Ohara, H. Yamamoto, J. Org. Chem. 1996, 61, 4196–4197.
[11] T. Maki, K. Ishihara, H. Yamamoto, Org. Lett. 2005, 7, 5043–5046.
[12] T. Maki, K. Ishihara, H. Yamamoto, Org. Lett. 2006, 8, 1431–1434.
[13] R. Yamashita, A. Sakakura, K. Ishihara, Org. Lett. 2013, 15, 3654–3657.
[14] N. Gernigon, R. M. Al-Zoubi, D. G. Hall, J. Org. Chem. 2012, 77, 8386–
8400.
[33] E. Calcio Gaudino, D. Carnaroglio, M. A. G. Nunes, L. Schmidt, E. M. M.
Flores, C. Deiana, Y. Sakhno, G. Martra, G. Cravotto, Catal. Sci. Technol.
2014, 4, 1395–1399.
[34] S. Nagarajan, P. Ran, P. Shanmugavelan, M. Sathishkumar, A. Ponnuswa-
my, K. S. Nahm, G. G. Kumar, New J. Chem. 2012, 36, 1312–1319.
[35] V. K. Das, R. R. Devi, P. K. Raul, A. J. Thakur, Green Chem. 2012, 14, 847–
854.
[36] S. Ghosh, A. Bhaumik, J. Mondal, A. Mallik, S. S. Bandyopadhyay, C. Mu-
khopadhyay, Green Chem. 2012, 14, 3220–3229.
[37] P. S. Chaudhari, S. D. Salim, R. V. Sawant, K. G. Akamanchi, Green Chem.
2010, 12, 1707–1710.
[38] K. Takahashi, M. Shibagaki, H. Kuno, H. Kawakami, H. Matsushita, Bull.
Chem. Soc. Jpn. 1989, 62, 1333–1334.
[39] B. M. Choudary, V. Bhaskar, M. L. Kantam, K. K. Rao, K. V. Raghavan,
Catal. Lett. 2001, 74, 207–211.
[40] K. V. N. S. Srinivas, B. Das, J. Org. Chem. 2003, 68, 1165–1167.
[41] M. Tamura, T. Tonomura, K. Shimizu, A. Satsuma, Green Chem. 2012, 14,
717–724.
[42] Md. A. Ali, S. M. A. Siddiki, K. Kon, K. Shimizu, ChemCatChem 2015, 7,
2705–2710.
[15] R. M. Al-Zoubi, O. Marion, D. G. Hall, Angew. Chem. Int. Ed. 2008, 47,
2876–2879; Angew. Chem. 2008, 120, 2918–2921.
[16] K. Arnold, A. S. Batsanov, B. Davies, A. Whiting, Green Chem. 2008, 10,
124–134.
[17] E. K. W. Tam, Rita, L. Y. Liu, A. Chen, Eur. J. Org. Chem. 2015, 1100–1107.
[18] R. M. Lanigan, P. Starkov, T. D. Sheppard, J. Org. Chem. 2013, 78, 4512–
4523.
´
[19] D. C. Lenstra, F. P. J. T. Rutjes, J. Mecinovic, Chem. Commun. 2014, 50,
5763–5766.
[20] H. Lundberg, F. Tinnis, H. Adolfsson, Chem. Eur. J. 2012, 18, 3822–3826.
[21] C. L. Allen, A. R. Chhatwal, J. M. J. Williams, Chem. Commun. 2012, 48,
666–668.
[22] F. Tinnis, H. Lundberg, H. Adolfsson, Adv. Synth. Catal. 2012, 354, 2531–
2536.
[43] M. A. Ali, S. M. A. H. Siddiki, K. Kon, J. Hasegawa, K. Shimizu, Chem. Eur.
J. 2014, 20, 14256–14260.
[44] H. Charville, D. A. Jackson, G. Hodges, A. Whiting, M. R. Wilson, Eur. J.
Org. Chem. 2011, 5981–5990.
[23] H. Lundberg, F. Tinnis, H. Adolfsson, Synlett 2012, 2201–2204.
[24] A. Chandra Shekhar, A. R. Kumar, G. Sathaiah, V. L. Paul, M. Sridhar, P. S.
Rao, Tetrahedron Lett. 2009, 50, 7099–7101.
[25] Y. Terada, N. Ieda, K. Komura, Y. Sugi, Synthesis 2008, 2318–2320.
[26] K. Steliou, M. A. Poupart, J. Am. Chem. Soc. 1983, 105, 7130–7138.
[27] J. W. Comerford, J. H. Clark, D. J. Macquarrie, S. W. Breeden, Chem.
Commun. 2009, 2562–2564.
[45] M. Tamura, K. Shimizu, A. Satsuma, Appl. Catal. A 2012, 433–434, 135–
145.
[46] K. Shimizu, T. Higuchi, E. Takasugi, T. Hatamachi, T. Kodama, A. Satsuma,
J. Mol. Catal. A 2008, 284, 89–96.
[28] N. Narender, P. Srinivasu, S. J. Kulkarni, K. V. Raghavan, Green Chem.
2000, 2, 104–105.
[29] K. V. V. Krishna Mohan, N. Narender, S. J. Kulkarni, Green Chem. 2006, 8,
368–372.
[47] T. Okuhara, Chem. Rev. 2002, 102, 3641–3666.
[48] S. Kobayashi, K. Manabe, Acc. Chem. Res. 2002, 35, 209–217.
[49] K. Nakajima, Y. Baba, R. Noma, M. Kitano, J. N. Kondo, S. Hayashi, M.
Hara, J. Am. Chem. Soc. 2011, 133, 4224–4227.
[30] A. Sakthivel, K. Komura, S. J. Huang, P. H. Wu, S. B. Liu, Y. Sasaki, Y. Sugi,
Ind. Eng. Chem. Res. 2010, 49, 65–71.
[31] K. Komura, Y. Kakano, M. Koketsu, Green Chem. 2011, 13, 828–831.
[32] M. Hosseini-Sarvari, H. Sharghi, J. Org. Chem. 2006, 71, 6652–6654.
Received: June 12, 2015
Revised: July 6, 2015
Published online on September 11, 2015
ChemCatChem 2015, 7, 3555 – 3561
www.chemcatchem.org
3561
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim