Heterogeneously catalyzed self-condensation of primary amines to secondary amines by supported copper catalysts

Article abstract of DOI:10.1039/c3cy00405h

Self-condensation of primary amines to symmetrically substituted secondary amines could efficiently be promoted by an inexpensive supported copper catalyst, Cu/Al₂O₃, easily prepared by the reduction of the hydroxide precursor, Cu(OH)_x/Al₂O₃. Various kinds of structurally diverse primary amines including benzylamine, picolylamine, and aliphatic amine derivatives could selectively be converted into the corresponding secondary amines in moderate to excellent yields without any cocatalysts such as bases and stabilizing ligands in 1 atm of Ar or H ₂. The reactions in H₂ showed higher selectivities to desired secondary amines than those in Ar. The roles of H₂ are the promotion of hydrogenation of N-alkylimines and the stabilization of active Cu(0) species. In addition, in the presence of Cu/Al₂O₃, unsymmetrically substituted secondary amines could efficiently be synthesized by N-alkylation of primary amines with alcohols and reductive amination of aldehydes. The observed catalysis was truly heterogeneous, and the retrieved Cu/Al₂O₃ catalyst could be reused for self-condensation without a significant loss of its catalytic performance. The reaction mechanism involving dehydrogenation of primary amines and condensation to N-alkylimines followed by hydrogenation, the so-called "borrowing hydrogen pathway", has been proposed. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3cy00405h

Catalysis
Science & Technology
View Article Online
View Journal
PAPER
Heterogeneously catalyzed self-condensation of
primary amines to secondary amines by supported
Cite this: DOI: 10.1039/c3cy00405h
copper catalysts†
Insu Kim, Shintaro Itagaki, Xiongjie Jin, Kazuya Yamaguchi and Noritaka Mizuno*
Self-condensation of primary amines to symmetrically substituted secondary amines could efficiently be
promoted by an inexpensive supported copper catalyst, Cu/Al2O3, easily prepared by the reduction of the
hydroxide precursor, Cu(OH)x/Al2O3. Various kinds of structurally diverse primary amines including
benzylamine, picolylamine, and aliphatic amine derivatives could selectively be converted into the
corresponding secondary amines in moderate to excellent yields without any cocatalysts such as bases and
stabilizing ligands in 1 atm of Ar or H2. The reactions in H2 showed higher selectivities to desired secondary
amines than those in Ar. The roles of H2 are the promotion of hydrogenation of N-alkylimines and the
stabilization of active Cu(0) species. In addition, in the presence of Cu/Al2O3, unsymmetrically substituted
secondary amines could efficiently be synthesized by N-alkylation of primary amines with alcohols and
reductive amination of aldehydes. The observed catalysis was truly heterogeneous, and the retrieved Cu/Al2O3
catalyst could be reused for self-condensation without a significant loss of its catalytic performance. The
reaction mechanism involving dehydrogenation of primary amines and condensation to N-alkylimines
followed by hydrogenation, the so-called ‘‘borrowing hydrogen pathway’’, has been proposed.
Received 12th June 2013,
Accepted 25th June 2013
DOI: 10.1039/c3cy00405h
www.rsc.org/catalysis
advantages of N-alkylation with alcohols and amines are
(i) high selectivity to desired secondary amines, (ii) formation
Introduction
Secondary amines have been utilized as useful synthons for
many profitable chemicals such as pharmaceuticals, agrochemicals,
and bioactive compounds.1 N-alkylation of primary amines with
alkyl halides (R0–X) in the presence of stoichiometric amounts of
inorganic or organic bases has generally been utilized for synthesis
of secondary amines [eqn (1)].1
of only water or NH3 as a byproduct, and (iii) use of readily
available, inexpensive, and less-toxic alcohols and amines as
alkylating reagents. To develop efficient N-alkylation systems
with alcohols and amines, the design of catalysts with both
dehydrogenation and hydrogenation abilities is required. To
date, many efficient catalysts including homogeneous as well as
heterogeneous ones for N-alkylation with alcohols have been
reported.3
(1)
Self-condensation of primary amines (N-alkylation with
primary amines) is also a very attractive procedure for synthesis
of symmetrically substituted secondary amines. However, self-
condensation of primary amines is much more difficult and
less reported than N-alkylation with alcohols. To date, there
have been several homogeneously catalyzed self-condensation
systems using platinum group metal-based complexes such
as PtCl2(PPh3)2,4 RuCl2(PPh3)3,5 and [Cp*IrCl2]2.6 Recently,
Yamashita and co-workers have reported the neopentyl-substituted
PNP–iridium complex/NaH-catalyzed highly efficient self-
condensation of aliphatic amines.7 However, these systems have
disadvantages of (i) indispensable use of cocatalysts, (ii) use
of expensive platinum group metal-based catalysts, (iii)
harsh reaction conditions (e.g., temperatures of 170–185 1C),
However, this classical stoichiometric procedure produces
large amounts of byproducts and frequently shows low selectivity
to desired secondary amines caused by undesirable over-alkylation
to tertiary amines and quaternary ammonium compounds.1
Recently, many alternative N-alkylation procedures for synthesis
of secondary amines with proper alkylating reagents such as
alcohols and amines have extensively been examined.2 The
Department of Applied Chemistry, School of Engineering, The University of Tokyo,
7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan. E-mail: tmizuno@mail.ecc.u-tokyo.ac.jp;
Fax: +81-3-5841-7220; Tel: +81-3-5841-7272
† Electronic supplementary information (ESI) available: Fig. S1 and S2. See DOI:
10.1039/c3cy00405h
c
This journal is The Royal Society of Chemistry 2013
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
(iv) difficulty in catalyst/product separation and reuse of catalysts, JNM-EX-270. 1H and 13C NMR spectra were measured at
and/or (v) limited scope of substrates. If widely applicable and 270 and 67.8 MHz, respectively, with TMS as an internal
efficient self-condensation systems using easily recoverable and standard (d = 0 ppm). ICP-AES analyses were performed with
reusable heterogeneous catalysts could be developed, they would be Shimadzu ICPS-8100. Thermogravimetric and differential
more desirable from the standpoint of green chemistry.8
thermal analyses (TG-DTA) were performed on Rigaku Thermo
With regard to heterogeneously catalyzed self-condensation of plus TG 8120. XRD patterns were measured with Rigaku
primary amines, there have been only four reports to date, to the Smartlab using Cu Ka radiation (40 kV–200 mA). X-Ray absorp-
best of our knowledge. Although supported platinum group metal tion spectra of Cu(OH)x/Al2O3, CuO, and Cu foil were recorded
catalysts such as Pt/C9 and Ru/C10 have been reported to be active at the NW10A beamline of PF at KEK, Japan (proposal no.
for aliphatic amines, they are not applicable to benzylic amines due 2007G096). The data were analyzed using REX2000 software
to the hydrogenolysis of substrates to toluene and benzene. In (version 2.5, Rigaku). Al2O3 (BET surface area: 160 m2 gꢀ1, Cꢀa1t.
addition, Pt/C and Ru/C require microwave irradiation and large No. KHS-24, Sumitomo Chemical), TiO2 (BET: 316 m2
g ,
amounts of Al powder (as a cocatalyst), respectively, to promote the Cat. No. ST-01, Ishihara Sangyo Kaisya), CeO2 (BET: 111 m2 gꢀ1
,
reaction efficiently.9,10 Pt–Sn/Al2O3 can catalyze self-condensation of Cat. No. 544841-25G, Aldrich), and SiO2 (BET: 342 m2 gꢀ1, Cat. No.
a wide range of primary amines including benzylic ones.11 Besides CARiACT Q-10, Fuji Silysia Chemical) were commercially available.
platinum group metal-based catalysts, Shi and co-workers have Metal salts, solvents, and substrates were obtained from Kanto
quite recently reported for the first time efficient self-condensation Chemical, TCI, Wako, or Aldrich (reagent grade), and purified prior
of primary amines with a non-noble metal-based catalyst (NiCu- to the use (if required).13
FeOx).12 The development of efficient non-noble metal-based
heterogeneous catalysts for self-condensation is still one of the
most challenging tasks for catalyst researchers.
Preparation and characterization of Cu/Al2O3
Cu/Al2O3 was prepared by reduction of the corresponding
hydroxide precursor, Cu(OH)x/Al2O3, using H2. It was con-
firmed by the TG-DTA analysis that Cu(OH)2 was completely
reduced to Cu(0) species in 1 atm of H2 at 175 1C (Fig. 1). In
contrast, the reduction of commonly utilized CuCl2ꢁ2H2O and
Cu(OAc)2ꢁH2O (OAc = acetate) to Cu(0) species required much
higher temperatures (500 1C for CuCl2ꢁ2H2O; 278 1C for
Cu(OAc)2ꢁH2O) than that of Cu(OH)2. For the above-mentioned
reasons, we chose the hydroxide as a precursor for Cu/Al2O3.
Therefore, Cu(OH)x/Al2O3 (with highly dispersed Cu(II) hydro-
xide species on Al2O3) was first prepared according to the
following procedure.14 An aqueous solution of CuCl2ꢁ2H2O
(8.3 mM, 60 mL) containing Al2O3 powder (2.0 g, calcined
at 550 1C) was vigorously stirred at room temperature. After
15 min, the pH of the solution was quickly adjusted to 12.0 by
addition of an aqueous solution of NaOH (1.0 M), and the
resulting slurry was further stirred for 24 h. The solid was then
filtered off, washed with a large amount of water, and dried
in vacuo to afford 2.0 g of Cu(OH)x/Al2O3 as a light blue powder.
The content of copper in Cu(OH)x/Al2O3 was 1.5 wt%. The XRD
Herein, we report that self-condensation of primary amines to
symmetrically substituted secondary amines can be promoted by
an easily prepared and inexpensive supported copper catalyst,
Cu/Al2O3, from the hydroxide precursor with highly dispersed
copper species. Various kinds of primary amines including
benzylamine, picolylamine, and aliphatic amine derivatives can
be converted into the corresponding symmetrically substituted
secondary amines in 1 atm of Ar or H2 [eqn (2)]. In addition,
Cu/Al2O3 can be applied to N-alkylation of primary amines with
alcohols [eqn (3)] and reductive amination of aldehydes with H2
[eqn (4)] to synthesize unsymmetrically substituted secondary
amines. The present system has the following significant advan-
tages; (i) applicability to various primary amines, (ii) high
selectivity to desired secondary amines (in particular in the
presence of H2), (iii) no use of cocatalysts, (iv) simple catalyst/
product separation and work-up, (v) reusability of the catalyst,
and (vi) use of an inexpensive copper-based catalyst.
(2)
(3)
(4)
Experimental section
Instruments and reagents
GC analyses were performed on Shimadzu GC-2014 with a FID
detector equipped with a capillary column (InertCap-1, I.D. =
0.25 mm, length = 60 m). GC-MS spectra were recorded on
Shimadzu GCMS-QP2010 at an ionization voltage of 70 eV.
Fig. 1 TG profile of Cu(OH)2 measured in 1 atm of H2 (progress temperature:
5 1C minꢀ1).
1
Liquid-state H and 13C NMR spectra were recorded on JEOL
c
Catal. Sci. Technol.
This journal is The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
pattern of Cu(OH)x/Al2O3 was the same as that of the parent successively placed Cu/Al2O3 (obtained by the reduction of the
Al2O3 support, and no signals due to copper oxides and hydroxides hydroxide precursor, 5 mol%), primary amine (0.5 mmol),
were observed (Fig. S1, ESI†). In the radial distribution function internal standard (typically biphenyl), and mesitylene (2 mL).
from the Fourier transformation of the k3-weighed EXAFS for Then, the resulting reaction mixture was stirred at 140 1C in
Cu(OH)x/Al2O3, the signals due to the Cu–O–Cu and Cu–Cu 1 atm of Ar or H2. The conversion and yield were periodically
shells were hardly observed (Fig. S2, ESI†).14a The absence of determined by GC analysis. After the reaction was completed,
these shell signals suggests that no networks of linked copper Cu/Al2O3 was separated by filtration and further washed
species were generated in Cu(OH)x/Al2O3. Thus, Cu(II) hydro- with mesitylene before reuse. As for isolation of products, an
xide species are highly dispersed on Al2O3. By the reduction of internal standard was not used. After primary amines were
Cu(OH)x/Al2O3 with H2 (1 atm) at 180 1C for 30 min, Cu/Al2O3 completely consumed, Cu/Al2O3 was filtered off, and then the
(dark red powder) was prepared just before the reaction. The XRD filtrate was concentrated by evaporation of mesitylene, giving
pattern of Cu/Al2O3 was the same as that of the parent Al2O3 the corresponding secondary amines. If required, the (crude)
support, and no signals due to Cu(0) species were observed, products were subjected to column chromatography on silica
while signals due to Cu(0) species were clearly observed in gel (Silica Gel 60N (63–210 mm), Kanto Chemical, 2.5 cm ID ꢂ
the case of a physical mixture of Cu metal (1 wt%) and Al2O3 15 cm length, eluent: a mixed solvent of diethyl ether and ethyl
(Fig. S1, ESI†). These results suggest that Cu(0) species are acetate (7 : 3, v/v) containing triethylamine (1 vol%)). The
highly dispersed on Al2O3. CuCl2/Al2O3 and Cu(OAc)2/Al2O3 identification of the products was carried out by NMR
were prepared according to the reported procedures.14 Other (1H and 13C) and/or GC-MS analyses (see the ESI†).
supported metal catalysts were also prepared according to the
similar procedure to that described above.
Results and discussion
Catalytic reactions
Effects of catalysts and solvents
All operations were performed in a glove box or using a Schlenk Initially, various kinds of supported metal catalysts were pre-
flask. A typical procedure for self-condensation of primary pared (copper, silver, gold, ruthenium, rhodium, palladium,
amines to secondary amines is as follows. Into a reactor were and platinum on Al2O3, M/Al2O3, see the Experimental section
Table 1 Self-condensation of 1a to 2a under various reaction conditions (optimization)a
Yield (%)
Entry
Catalyst
Solvent
Time (h)
2a
3a
17
3
11
16
6
o1
14
10
o1
25
2
3
2
3
5
1
9
4
4
4a
5a
1
2
3
Cu/Al2O3
Cu/Al2O3
Cu/Al2O3
Cu/Al2O3
Cu/Al2O3
Cu/Al2O3
Cu/TiO2
Cu/CeO2
Mesitylene
Mesitylene
Diglyme
o-Dichlorobenzene
DMSO
3
24
3
3
3
3
3
3
3
24
24
24
24
24
24
24
24
24
24
3
64
97
31
o1
o1
o1
38
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
11
o1
o1
o1
o1
35
4
5b
6c
DMF
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
o1
24
7
8
9
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
Mesitylene
38
Cu/SiO2
Cu(OH)x/Al2O3
CuCl2/Al2O3 (pretreated with H2 at 180 1C)
Cu(OAc)2/Al2O3 (pretreated with H2 at 180 1C)
o1
10
11
12
13d
14
15
16
17
18
19
20
21
22
23
24
25
26d
27
47
o1
o1
o1
o1
o1
o1
o1
o1
o1
2
Cu metal + Al2O3
CuO
Cu(OH)2
Cu2O
CuCl2ꢁ2H2O
Cu(OAc)2ꢁH2O
CuCl
Ag/Al2O3
Au/Al2O3
Ru/Al2O3
Rh/Al2O3
Pd/Al2O3
Pt/Al2O3
Al2O3
o1
3
3
3
3
3
3
24
1
29
3
28
o1
16
21
o1
o1
o1
58
47
27
2
18
7
29
o1
o1
o1
None
o1
o1
o1
a
Reaction conditions: catalyst (metal: 5 mol%), 1a (0.5 mmol), solvent (2 mL), 140 1C, in 1 atm of Ar. Yields were determined by GC.
N,N-Dimethylbenzylamine was formed as a byproduct (52%). N-Benzylformamide was formed as a byproduct (51%). Al2O3 (100 mg).
b
c
d
c
This journal is The Royal Society of Chemistry 2013
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
for the preparation), and their catalytic activities and selectivities
The selection of solvents was also very important. Mesitylene
were assessed by the self-condensation of benzylamine (1a) to was the most suitable solvent for the self-condensation
dibenzylamine (2a) at 140 1C in 1 atm of Ar. Under the conditions (Table 1, entry 1). Diethylene glycol dimethyl ether (diglyme)
described in Table 1, no reaction proceeded in the absence of the gave moderate yields of 2a and 3a (Table 1, entry 3). o-Dichloro-
catalysts (Table 1, entry 27) or the presence of just Al2O3 (Table 1, benzene afforded only 3a (Table 1, entry 4). Under the present
entry 26). Among various supported metal catalysts examined, reaction conditions, dimethyl sulfoxide (DMSO) and N,N-dimethyl-
Cu/Al2O3 showed the highest catalytic performance, and the formamide (DMF) readily decomposed and acted as methylating
reaction for 3 h gave 2a and N-benzylidenebenzylamine (3a) and formylating reagents. Thus, the reaction of 1a in DMSO and
in 64% and 17% yields, respectively (Table 1, entry 1). In this DMF afforded N,N-dimethylbenzylamine (52% yield) and N-benzyl-
case, over-alkylation, over-dehydrogenation, hydrogenolysis, formamide (41% yield), respectively, as major products (Table 1,
and hydrogenation of the aromatic ring did not proceed at entries 5 and 6).
all. Although Ru/Al2O3 gave moderate yields of 2a and 3a, an
Therefore, Cu/Al2O3 was the best catalyst, and mesitylene
over-dehydrogenation product of benzonitrile (4a) was undesirably was the best solvent; when the self-condensation of 1a was
formed (Table 1, entry 22). It has been reported that hydro- carried out with Cu/Al2O3 in mesitylene for 24 h, the desired
genolysis of benzylamine proceeds to form toluene (5a) and secondary amine 2a was obtained in 97% yield with a small
benzene in the presence of Pd/C.15 Similarly, when using amount of 3a (Table 1, entry 2). We hereafter carried out all
Rh/Al2O3, Pd/Al2O3, and Pt/Al2O3, the hydrogenolysis proceeded reactions using Cu/Al2O3 in mesitylene.
to some extent, resulting in formation of 5a as a byproduct
Heterogeneous catalysis and catalyst reuse
(Table 1, entries 23–25). In addition, over-dehydrogenation
proceeded with Rh/Al2O3, Pd/Al2O3, and Pt/Al2O3 (Table 1, It is important to verify that the observed catalysis is derived
entries 23–25). In the case of Ag/Al2O3 and Au/Al2O3, almost from solid Cu/Al2O3 or leached copper species. Therefore,
no conversion of 1a was observed likely due to their poor the following control experiments were carried out. The self-
dehydrogenation and hydrogenation abilities (Table 1, entries 20 condensation of 1a to 2a was carried out under the conditions
and 21).16 The effect of supports was significant. Cu/Al2O3 showed described in Table 1, and the Cu/Al2O3 catalyst was removed
higher catalytic performance than Cu/TiO2 and Cu/CeO2 (Table 1, from the reaction mixture by filtration at approximately 50%
entries 1, 7, and 8). The reaction hardly proceeded with Cu/SiO2 conversion of 1a. After removal of the catalyst, the filtrate was
(Table 1, entry 9). It is likely that the acidic nature of SiO2 inhibits again heated at 140 1C. Although ca. 0.2% copper species
the abstraction of an amine hydrogen.
(based on the catalyst used) were detected in the filtrate, no
Next, the self-condensation of 1a was carried out with further production of 2a was observed. These results can rule
various copper-based catalysts. The reaction hardly proceeded out any contribution to the observed catalysis from copper
with simple copper salts of CuCl2ꢁ2H2O, Cu(OAc)2ꢁH2O, and species that leached into the reaction solution, and the
CuCl (Table 1, entries 17–19). Bulk copper oxides and hydroxide observed catalysis is intrinsically heterogeneous.17 In addition,
such as CuO, Cu(OH)2, and Cu2O were also not effective for the the Cu/Al2O3 catalyst could be reused without a significant loss
self-condensation (Table 1, entries 14–16). The catalytic activity of its high catalytic performance (Table 2, entries 3 and 4).
of a physical mixture of Cu metal powder and Al2O3 was much
Synthetic scope
lower than that of Cu/Al2O3 (Table 1, entry 13). The hydroxide
precursor, Cu(OH)x/Al2O3, showed moderate catalytic performance, The scope of the present Cu/Al2O3-catalyzed self-condensation
affording 2a and 3a in 47% and 25% yields, respectively, for 24 h of primary amines to secondary amines was next investigated.
(Table 1, entry 10). We confirmed that the reaction with Cu(OH)x/ As shown in Table 2, various kinds of structurally diverse
Al2O3 possessed an induction period under the present conditions primary amines could be converted into the corresponding
and that Cu(II) species in Cu(OH)x/Al2O3 were reduced to Cu(0) symmetrically substituted secondary amines in 1 atm of Ar or
species during the period. All the above-mentioned results show H2. In all cases, the reactions in H2 showed higher selectivities
that highly dispersed Cu(0) species would play an important role in to desired secondary amines than those in Ar. The roles of H2
the present self-condensation of primary amines.
are the promotion of hydrogenation of N-alkylimine intermedi-
Although we also utilized CuCl2/Al2O3 and Cu(OAc)2/Al2O3 as ates and the stabilization of active Cu(0) species. Benzylamine
catalysts14 for the self-condensation of 1a (they were pretreated (1a) as well as its derivatives with electron-donating (1b, 1d, and
with H2 at 180 1C before the reaction), their blue colour did not 1e) and slightly electron-withdrawing (1c) substituents on the
change during the H2 treatment as well as during the reaction, phenyl rings efficiently proceeded in both Ar and H2, giving the
and no conversion of 1a was observed in both cases (Table 1, corresponding dibenzylamine derivatives in moderate to high yields
entries 11 and 12). As mentioned in the Experimental section, (Table 2, entries 1–12). Although the reaction of p-trifluoromethyl-
we confirmed by the TG-DTA analyses that the reduction benzylamine (1f) containing the strongly electron-withdrawing CF3
of CuCl2ꢁ2H2O and Cu(OAc)2ꢁH2O to Cu(0) species hardly group in Ar gave the corresponding secondary amine in only 8%
proceeded at 180 1C. In contrast, the hydroxide species yield (Table 2, entry 13), the yield dramatically increased up to
could easily be reduced. Therefore, Cu(OH)x/Al2O3 is a suitable 83% when carrying out the reaction in H2 (Table 2, entry 14). The
precursor for preparation of the active Cu/Al2O3 catalyst under present system could be applied to 2-, 3-, and 4-picolylamines
relatively milder conditions.
(1g–1i), affording the corresponding dipicolylamine derivatives
c
Catal. Sci. Technol.
This journal is The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
Table 2 Self-condensation of various primary aminesa
Yield (%)
Time
Entry
Substrate
(h)
2
3
1
Ar
24
24
24
24
97 (81)
98
98
3
o1
o1
3
2
H2
H2
H2
1a
3b
4b
83
Scheme 1 Cu/Al2O3-catalyzed N-alkylation of primary amines with alcohols.
Reaction conditions: Cu/Al2O3 (5 mol%), primary amine (0.5 mmol), alcohol
(2 mmol), mesitylene (2 mL), 140 1C, in 1 atm of Ar, 1 h.
5
6
Ar
H2
15
15
65
89 (93)
11
1
1b
1c
7
8
Ar
H2
24
24
67
83
6
2
Besides synthesis of symmetrically substituted secondary
amines demonstrated in Table 2, unsymmetrically substituted
secondary amines could successfully be synthesized in the
presence of Cu/Al2O3 (Schemes 1 and 2). In these cases, over-
alkylated products were hardly formed; a small amount of the
corresponding dialkylated product was formed in the synthesis
of 2m (below 3%), and no over-alkylated products were formed
for 2l and 2n. When the reaction of primary amines with
alcohols as alkylating reagents (four equivalents)‡ was carried out
in the presence of Cu/Al2O3, the corresponding unsymmetrically
substituted secondary amines could selectively be obtained in
high yields (Scheme 1). In addition, Cu/Al2O3 could catalyze the
reductive amination of benzaldehyde with aniline in 1 atm of H2,
giving N-phenylbenzylamine in a quantitative yield (Scheme 2).
Platinum group metal-based catalysts have frequently been utilized
for reductive amination with H2,19 and to date there have been only
two reports on reductive amination with H2 using copper-based
catalysts,20 as far as we know.
9
10
Ar
H2
24
24
69
79 (93)
3
o1
1d
1e
1f
11
12
Ar
H2
15
15
80
95 (94)
3
1
13
14
Ar
H2
24
24
8
83
39
1
15
16
Ar
H2
24
24
64
74 (57)
2
o1
1g
1h
1i
17
18
Ar
H2
24
24
47
88
10
1
19
20
Ar
H2
24
24
72
78
o1
o1
Possible reaction pathway
21
22
23
24
CH3(CH2)7NH2
CH3(CH2)11NH2
1j
Ar
H2
Ar
H2
24
24
24
24
69
90 (93)
50
3
o1
3
o1
The reaction profiles for the present Cu/Al2O3-catalyzed self-
condensation of primary amines to secondary amines showed
that the corresponding N-alkylimines were initially produced
followed by the formation of the corresponding secondary
amines. During the transformation, aldehydes were detected
by GC-MS analyses albeit in only small amounts (typically below
1%). Aldehydes were likely produced by the hydrolytic decom-
position (with water containing Cu/Al2O3 and/or the solvent) of
the aldimine intermediates formed through the dehydrogena-
tion of the starting primary amines. The mass analyses of the
gas phase for the self-condensation of 1a showed the formation
of H2 and NH3 during the transformation. As above-mentioned,
Cu/Al2O3 showed the high catalytic performance for reductive
amination with H2. In addition, it was confirmed in a separate
experiment that the hydrogenation of 3a to 2a in 1 atm of H2
efficiently proceeded in the presence of Cu/Al2O3 (Scheme 3).
Experimental evidence indicates that the present Cu/Al2O3-
catalyzed self-condensation of primary amines to secondary
amines possibly proceeds through the following sequential
1k
81 (95)
a
Reaction conditions: Cu/Al2O3 (Cu:
5
mol%), primary amine
(0.5 mmol), mesitylene (2 mL), 140 1C, in 1 atm of Ar or H2. Yields
were determined by GC. The values in the parentheses are isolated
yields after complete conversions of 1 and/or hydrogenations of 3
(24–60 h, see the ESI). Reuse experiments (entry 3 for the 1st reuse;
entry 4 for the 2nd reuse).
b
(Table 2, entries 15–20). In particular, the self-condensation of 1i
is an attractive alternative for the synthesis of 2,20-dipicolylamine
(Table 2, entries 19 and 20). 2,20-Dipicolylamine has been
utilized as a ligand as well as a synthon for several fluorescent
probes and generally synthesized by reductive amination using
NaBH4.18 Although efficient heterogeneous catalysts such as
Pt–Sn/Al2O3 (ref. 11) and NiCuFeOx (ref. 12) have been reported
for N-alkylation (including self-condensation of primary amines),
their applicabilities to synthesis of dipicolylamine derivatives have
never been mentioned. Moreover, the present Cu/Al2O3-catalyzed
system was also applicable to aliphatic primary amines (1j and 1k)
to synthesize dialkylamines (Table 2, entries 21–24).
‡ We confirmed in the separate experiments that Cu/Al2O3 could efficiently
promote dehydrogenation of alcohols.21
c
This journal is The Royal Society of Chemistry 2013
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Conclusions
We have successfully developed an efficient procedure for
synthesis of symmetrically substituted secondary amines via
the Cu/Al2O3-catalyzed self-condensation of primary amines in
1 atm of Ar or H2. This system was applicable to various kinds
of structurally diverse primary amines including benzylamine,
picolylamine, and aliphatic amine derivatives as starting materials.
In addition, unsymmetrically substituted secondary amines could
also be synthesized by N-alkylation of primary amines with alcohols
and reductive amination of aldehydes with H2. The observed
catalysis for self-condensation was truly heterogeneous in nature,
the rather inexpensive copper-based catalyst (in comparison with
platinum group metal-based catalysts) could be used, and Cu/Al2O3
could be reused without an appreciable loss of its high perfor-
mance. The present reactions proceeded through the typical
borrowing hydrogen pathway, and the high catalytic perfor-
mance of Cu/Al2O3 is derived from its excellent ability for
dehydrogenation and hydrogenation. This convenient procedure
with the easily prepared, inexpensive Cu/Al2O3 catalyst demon-
strated herein will provide a new green route to secondary
amines starting from various substrates such as primary amines,
alcohols, and aldehydes, which can completely avoid utilization
of stoichiometric reagents (e.g., alkyl halides) and cocatalysts
(e.g., bases and stabilizing ligands) and formation of inorganic
byproducts, and we hope that this system will be one of the
choices for synthesis of secondary amines.
Scheme 2 Cu/Al2O3-catalyzed reductive amination. Reaction conditions:
Cu/Al2O3 (5 mol%), aniline (0.5 mmol), benzaldehyde (0.55 mmol), mesitylene
(2 mL), 140 1C, in 1 atm of H2, 5 h.
Scheme 3 Cu/Al2O3-catalyzed hydrogenation of 3a. Reaction conditions: Cu/
Al2O3 (5 mol%), 3a (0.5 mmol), mesitylene (2 mL), 140 1C, in 1 atm of H2, 2 h.
reactions, the so-called ‘‘borrowing hydrogen pathway’’
(Scheme 4). Firstly, the dehydrogenation of a primary amine
to an aldimine takes place with the transitory formation of H2
(step 1 in Scheme 4). Then, the aldimine readily reacts with the
starting primary amine to generate the corresponding N-alkyli-
mine and NH3 (step 2 in Scheme 4). Even if the hydrolytic
decomposition of the aldimine to an aldehyde proceeds, this is
not a serious problem because the aldehyde formed can also
react with a primary amine to generate the corresponding
N-alkylimine (step 20 in Scheme 4). Finally, the hydrogenation
of the N-alkylimine proceeds to afford the desired secondary
amine (step 3 in Scheme 4). N-Alkylation with alcohols also
proceeds through this borrowing hydrogen pathway except that
the first step is alcohol dehydrogenation.‡
Acknowledgements
This work was supported in part by the Grants-in-Aid for
Scientific Research (B) from Ministry of Education, Culture,
Sports, Science and Technology (Grant Number 22360334).
Notes and references
1 (a) R. N. Salvatore, C. H. Yoon and K. W. Jung, Tetrahedron,
2001, 57, 7785; (b) M. B. Smith and J. March, March’s
Advanced Organic Chemistry: Reactions, Mechanisms, and
Structure, Wiley, Hodoken, New Jersey, 6th edn, 2007.
´
2 (a) G. Guillena, D. J. Ramon and M. Yus, Chem. Rev., 2010,
110, 1611; (b) T. D. Nixon, M. K. Whittlesey and J. M. J.
Williams, Dalton Trans., 2009, 753.
3 (a) S. P. Pathare and K. G. Akamanchi, Appl. Catal., A, 2013,
¨
452, 29; (b) A. Wetzel, S. Wockel, M. Schelwies, M. K. Brinks,
F. Rominger, P. Hofmann and M. Limbach, Org. Lett., 2013,
15, 266; (c) H. Liu, G. Chuah and S. Jaenicke, J. Catal., 2012,
292, 130; (d) J. Sun, X. Jin, F. Zhang, W. Hu, J. Liu and R. Li,
Catal. Commun., 2012, 24, 30; (e) R. Kawahara, K. Fujita and
R. Yamaguchi, Adv. Synth. Catal., 2011, 353, 1161;
( f ) K. Shimizu, K. Shimura, M. Nishimura and A.
Satsuma, RSC Adv., 2011, 1, 1310; (g) J. He, K. Yamaguchi
and N. Mizuno, Chem. Lett., 2010, 1182; (h) A. Corma,
T. Rodenas and M. J. Sabater, Chem.–Eur. J., 2010, 16, 254;
(i) L. He, X. Lou, J. Ni, Y. Liu, Y. Cao, H. He and K. Fan,
Scheme 4 The possible reaction pathway for the present Cu/Al2O3-catalyzed
self-condensation of primary amines to secondary amines.
c
Catal. Sci. Technol.
This journal is The Royal Society of Chemistry 2013

View Article Online
Catalysis Science & Technology
Paper
Chem.–Eur. J., 2010, 16, 13965; ( j) J. Kim, K. Yamaguchi and
N. Mizuno, J. Catal., 2009, 263, 205.
A. K. L. Yuen, A. J. Ward, A. F. Masters and
T. Maschmeyer, ChemCatChem, 2011, 3, 1496.
4 Y. Tsuji, J. Shida, R. Takeuchi and Y. Watanabe, Chem. Lett., 16 T. Bligaard, J. K. Nørslov, S. Dahl, J. Matthiese,
1984, 889. C. H. Christensen and J. Sehested, J. Catal., 2004, 224, 206.
5 B.-T. Khai, C. Concilio and G. Porzi, J. Organomet. Chem., 17 R. A. Sheldon, M. Wallau, I. W. C. E. Arends and
1981, 208, 249. U. Schuchardt, Acc. Chem. Res., 1998, 31, 485.
6 L. L. R. Lorentz-Petersen, P. Jensen and R. Madsen, Synth- 18 (a) H. Woo, S. Cho, Y. Han, W. Chae, D. Ahn, Y. You and
esis, 2009, 4110.
7 M. Yamashita, Y. Moroe, T. Yano and K. Nozaki, Inorg.
Chim. Acta, 2011, 369, 15.
W. Nam, J. Am. Chem. Soc., 2013, 135, 4771; (b) U. Reddy,
G. P. Das, S. Saha, M. Baidya, S. K. Ghosh and A. Das, Chem.
Commun., 2013, 49, 255; (c) S. Sumiya, Y. Shiraishi and
T. Hirai, J. Phys. Chem. A, 2013, 117, 1474; (d) A. S. Rao,
S. Singha, W. Choi and K. Ahn, Org. Biomol. Chem., 2012,
10, 8410; (e) M. A. R. Raycroft, C. I. Maxwell, R. A.
A. Oldham, A. S. Andrea, A. A. Neverov and R. S. Brown,
Inorg. Chem., 2012, 51, 10325; ( f ) S. Kim, C. Saracini,
M. A. Siegler, N. Drichko and K. D. Karlin, Inorg. Chem.,
2012, 51, 12603; (g) S. Mundinger, U. Jakob, P. Bichovski
and W. Bannwarth, J. Org. Chem., 2012, 77, 8968;
(h) B. A. Smith, W. J. Akers, W. M. Leevy, A. J. Lampkins,
S. Xiao, W. Wolter, M. A. Suckow, S. Achilefu and
B. D. Smith, J. Am. Chem. Soc., 2010, 132, 67;
(i) K. Komatsu, Y. Urano, H. Kojima and T. Nagano, J. Am.
Chem. Soc., 2007, 129, 13447.
8 (a) P. T. Anastas and J. C. Warner, Green Chemistry: Theory
and Practice, Oxford University Press, London, 1998;
(b) R. A. Sheldon, Green Chem., 2000, 2, G1;
(c) P. T. Anastas, L. B. Bartlett, M. M. Kirchhoff and
T. C. Williamson, Catal. Today, 2000, 55, 11;
(d) R. A. Sheldon and H. van Bekkum, Fine Chemical through
Heterogeneous Catalysis, Wiley, Weinheim, 2001.
9 A. Miyazawa, K. Saitou, K. Tanaka, T. M. Gadda, M. Tashiro,
G. K. S. Prakash and S. A. Olah, Tetrahedron Lett., 2006, 47, 1437.
10 A. M. Simion, T. Arimura, A. Miyazawa, C. Simion, G. K.
S. Prakash, G. A. Olah and M. Tashiro, Synth. Commun.,
2009, 39, 2859.
11 W. He, L. Wang, C. Sun, K. Wu, S. He, J. Chen, P. Wu and
Z. Yu, Chem.–Eur. J., 2011, 17, 13308.
12 X. Cui, X. Dai, Y. Deng and F. Shi, Chem.–Eur. J., 2013,
19, 3665.
13 Purification of Laboratory Chemicals, ed. D. D. Perrin and
W. L. F. Armarego, Pergamon Press, Oxford, UK, 3rd edn,
1988.
19 (a) R. L. Augustine, Heterogeneous Catalysis for the Synthetic
Chemist, Marcel Dekker, New York, 1996, p. 499;
(b) S. Ichikawa, T. Seki and T. Ikariya, Chem. Lett., 2012,
1628; (c) S. Wei, Z. Dong, Z. Ma, J. Sun and J. Ma, Catal.
´
Commun., 2013, 30, 40; (d) R. Cano, M. Yus and D. J. Ramon,
Tetrahedron, 2011, 67, 8079; (e) M. E. Domine,
´
´
14 (a) T. Katayama, K. Kamata, K. Yamaguchi and N. Mizuno,
ChemSusChem, 2009, 2, 59; (b) T. Oishi, T. Katayama,
M. C. Hernandez-Soto and Y. Perez, Catal. Today, 2011,
159, 2.
K. Yamaguchi and N. Mizuno, Chem.–Eur. J., 2009, 20 (a) S. Werkmeister, K. Junge and M. Beller, Green Chem.,
15, 7539; (c) K. Yamaguchi, T. Oishi, T. Katayama and
N. Mizuno, Chem.–Eur. J., 2009, 15, 10464.
2012, 14, 2371; (b) F. Santoro, R. Psaro, N. Ravasio and
F. Zaccheria, ChemCatChem, 2012, 4, 1249.
15 (a) A. J. Yap, A. F. Masters and T. Maschmeyer, 21 T. Mitsudome, Y. Mikami, K. Ebata, T. Mizugaki,
ChemCatChem, 2012, 4, 1179; (b) A. J. Yap, B. Chan,
K. Jitsukawa and K. Kaneda, Chem. Commun., 2008, 4804.
c
This journal is The Royal Society of Chemistry 2013
Catal. Sci. Technol.