Selective Activation of Alcohols in the Presence of Reactive Amines over Intermetallic PdZn: Efficient Catalysis for Alcohol-Based N-Alkylation of Various Amines

Article abstract of DOI:10.1021/acscatal.6b01677

Pd-based intermetallic compounds supported on Al₂O₃ (Pd_xM_y/Al₂O₃, where M = Bi, Cu, Fe, Ga, In, Pb, Sn, or Zn) were prepared and tested as catalysts for the selective activation of alcohols in the presence of reactive amines, which is highly challenging and is the key strategy for alcohol-based N-alkylation of amines. Although the Pd/Al₂O₃ catalyst exhibited a high catalytic activity, undesired side reactions such as amine dimerization (via amine activation) and C-O bond scission occurred, resulting in a poor yield of the N-alkylation product. In contrast, the PdZn/Al₂O₃ catalyst acted as an efficient catalyst for this reaction, displaying high catalytic activities, selectivities, and atom efficiencies and a wide substrate scope. Detailed kinetic and computational studies revealed that the relative affinity of Pd for alcohol and amine drastically changes by the formation of a PdZn intermetallic phase. On monometallic Pd, the adsorption and activation of amines are preferred over those of alcohols in terms of thermodynamic and kinetic aspects, respectively. However, this trend is inverted on PdZn, allowing preferential adsorption and activation of alcohols and, hence, selective N-alkylation.

Full text of DOI:10.1021/acscatal.6b01677

Research Article
pubs.acs.org/acscatalysis
Selective Activation of Alcohols in the Presence of Reactive Amines
over Intermetallic PdZn: Efficient Catalysis for Alcohol-Based
N‑Alkylation of Various Amines
,†
Shinya Furukawa,* Ryohei Suzuki, and Takayuki Komatsu*
Department of Chemistry, School of Science, Tokyo Institute of Technology, 2-12-1-E1-10 Ookayama, Meguro-ku, Tokyo 152-8551,
Japan
*
S Supporting Information
ABSTRACT: Pd-based intermetallic compounds supported on Al O3
2
(
Pd M /Al O , where M = Bi, Cu, Fe, Ga, In, Pb, Sn, or Zn) were
x y 2 3
prepared and tested as catalysts for the selective activation of alcohols in
the presence of reactive amines, which is highly challenging and is the key
strategy for alcohol-based N-alkylation of amines. Although the Pd/Al O3
2
catalyst exhibited a high catalytic activity, undesired side reactions such as
amine dimerization (via amine activation) and C−O bond scission
occurred, resulting in a poor yield of the N-alkylation product. In contrast,
the PdZn/Al O catalyst acted as an efficient catalyst for this reaction,
2
3
displaying high catalytic activities, selectivities, and atom efficiencies and a
wide substrate scope. Detailed kinetic and computational studies revealed
that the relative affinity of Pd for alcohol and amine drastically changes by the formation of a PdZn intermetallic phase. On
monometallic Pd, the adsorption and activation of amines are preferred over those of alcohols in terms of thermodynamic and
kinetic aspects, respectively. However, this trend is inverted on PdZn, allowing preferential adsorption and activation of alcohols
and, hence, selective N-alkylation.
KEYWORDS: N-alkylation, amine, alcohol, PdZn, intermetallic, catalyst
INTRODUCTION
cleavage), which eventually decrease the product yield. This
drawback has restricted the substrate scope of amines to less
reactive α-hydrogen-free amines such as aniline and its
derivatives. For other reactive amines, an excess of amine or
alcohol substrate is frequently used to obtain high product
yields. However, a stoichiometric feed of alcohol and amine is
highly desired for atom efficiency. To achieve efficient catalysis
for this reaction with a wide substrate scope, a specific reaction
system that activates alcohols, preferentially to amines, is
required. In homogeneous catalysis, much effort has been
■
N-Alkylamines are important building blocks comprising a
variety of functional organic molecules such as polymers,
pharmaceuticals, herbicides, and bioactive molecules. They are
typically synthesized by N-alkylation of lower amines, a key
synthetic methodology for C−N bond formation. Conventional
methods for N-alkylation of amines rely on the use of alkylating
agents such as aldehydes (in reductive amination) or alkyl
halides. However, the former requires a high H pressure or
reducing agents such as hydrides, and the latter yields a
1
2
11
1
devoted to developing efficient catalysts, and an excellent Fe-
stoichiometric amount of halogen waste. Thus, more efficient,
environmentally friendly protocols are required for N-alkylation
of amines.
based catalyst has recently been reported by Feringa and co-
9
,10
workers.
However, developing such a system in heteroge-
neous catalysis is highly challenging, and no promising catalyst
design has been reported in the literature. To overcome this
challenge, a drastic change in the chemical properties of active
metals may be needed.
Promising candidates for catalyst materials that provide such
well-modified reaction sites are intermetallic compounds.
Unlike solid−solution alloys, intermetallic compounds typically
comprise metal elements with significantly different chemical
properties and display well-defined bimetallic surfaces originat-
Recently, alcohol-based N-alkylation of amines has been
favored as a highly efficient synthetic methodology for
producing N-alkylamines. This reaction proceeds via a
combination of alcohol dehydrogenation and subsequent
reductive amination, where hydrogen transfer from the alcohol
to the intermediate imine is catalyzed by a transition metal,
2
3
4
5
6
7
8
9,10
such as Au, Ag, Cu, Pt, Pd, Ni, Ir, Rh, Ru, or Fe.
Alcohol acts as both an alkylating reagent and a hydrogen
source, thereby providing high atom efficiency, environmental
benignity, and easy handling under atmospheric pressure.
However, amines, in general, are more reactive than alcohols;
this leads to the frequent occurrence of undesired side reactions
such as amine dimerization and fragmentation (C−N bond
Received: June 14, 2016
Revised: July 25, 2016
©
XXXX American Chemical Society
5946
DOI: 10.1021/acscatal.6b01677
ACS Catal. 2016, 6, 5946−5953

ACS Catalysis
Research Article
ing from their specific crystal structures. An appropriate
combination of two metal elements provides not only a
substantially modified electronic structure of the active metal
component but also a uniform reaction environment. Indeed,
we have reported that some Pd-based intermetallic compounds
are highly active and act as selective catalysts for various organic
amine-α,α-d , CDN Isotopes, 99.3 atom % D, and benzyl
2
alcohol-α,α-d , ISOTEC, 98 atom % D) were used for
2
estimation of the kinetic isotope effect on α-C−H bond
activation. For recycling tests, the catalyst after reaction was
washed by decantation in the reactor with p-xylene three times
and finally once with n-hexane. The n-hexane rinse was
completely removed by decantation and subsequent drying
12,13
reactions, such as amine oxidation
hydrogenation of nitrostyrenes, where monometallic Pd
catalysts exhibit poor catalytic performances.
and chemoselective
14
−1
under an Ar flow (20 mL min ) at room temperature. The
washed catalyst was reused after the pretreatment step outlined
above.
In the study presented here, we prepared a series of Pd-based
intermetallic compounds supported on Al O (Pd M /Al O ,
Characterizations. The crystal structures of the catalysts
were determined by powder X-ray diffraction (XRD) with a
Rigaku RINT2400 diffractometer using a Cu Kα X-ray source.
Differential XRD patterns were obtained by subtracting the
pattern for the Al O support from that of the supported
2
3
x
y
2
3
where M = Bi, Cu, Fe, Ga, In, Pb, Sn, or Zn) and examined
their catalytic performances in alcohol-based N-alkylation of
various amines. The observed catalytic performances were
rationalized on an atomic level on the basis of kinetic and
computational studies. Herein, we report a novel and highly
efficient catalytic system for N-alkylation of Pd-based
intermetallic compounds.
2
3
catalysts. Transmission electron microscopy (TEM) was
conducted using a JEOL JEM-2010F microscope at an
accelerating voltage of 200 kV. To prepare the TEM specimen,
all samples were sonicated in tetrachloromethane and then
dispersed on a Cu grid supported by an ultrathin carbon film.
CO pulse chemisorption was performed to estimate the Pd
dispersion in the prepared catalysts. Prior to chemisorption, the
EXPERIMENTAL SECTION
Catalyst Preparation. Pd-based catalysts (Pd M /Al O ;
■
x
y
2
3
Pd loading, 3.0 wt %; M = Bi, Cu, Fe, Ga, In, Pb, Sn, or Zn)
were prepared by the successive impregnation (where M = Cu,
Ga, In, Pb, or Sn) or co-impregnation (where M = Bi, Fe, or
Zn) method using Al O (JRC-ALO-8, γ + θ phase; S = 148
−1
catalyst was pretreated under a H flow (60 mL min ) for 0.5
2
h. After the reduction pretreatment, He was introduced at the
same temperature for 30 min to remove the chemisorbed
hydrogen, followed by cooling to room temperature. A 5%
CO/He pulse was introduced into the reactor, and the passed
supplied CO was quantified downstream by a thermal
conductivity detector. Inductively coupled plasma atomic
emission spectroscopy (ICP-AES) analysis of the catalysts
was performed using a Shimadzu ICPE-9000 spectrometer. The
catalysts were dissolved using aqua regia and a HF solution.
Computational Details. Periodic density functional theory
calculations for surface reactions were performed using the
2
3
BET
2
−1
m g ) as a catalyst support. A specific amount of an aqueous
solution containing (NH ) PdCl alone or both (NH ) PdCl
4
2
4
4 2
4
and a second metal salt [typically, a nitrate except (NH )SnCl ]
4
6
in a 1:1 molar ratio (2:1 for PdGa) was added dropwise to a
vigorously stirred aqueous slurry of Al O (15 mL of ion-
2
3
exchanged water/g of Al O ). After being stirred for 15 min
2
3
under an air atmosphere, the slurry was completely dried on a
hot plate and ground to a fine powder. The sample was then
−1
reduced in a quartz tube under a hydrogen flow (60 mL min )
at a predetermined temperature for 1 h (Pd, 500 °C; Pd−Bi,
15
CASTEP code with Vanderbilt-type ultrasoft pseudopoten-
16
6
8
00 °C; Pd−Cu, Pd−Ga, Pd−In, Pd−Pb, Pd−Sn, and Pd−Zn,
00 °C; Fe, 900 °C). For successive impregnation, this
tials and a revised Perdew−Burke−Ernzerhof exchange-
17,18
correlation functional
based on the generalized gradient
procedure was repeated using Pd/Al O3 instead of bare
approximation. The plane-wave basis set was truncated at a
kinetic energy of 370 eV, and a Fermi smearing of 0.1 eV was
utilized. In this study, we chose the Pd(111), PdCu(110), and
PdZn(111) planes as the most stable surfaces. Monkhorst−
2
Al O . For the preparation of PdBi/Al O , a nitric acid solution
2
3
2
3
of Bi(NO ) ·5H O was added dropwise to a vigorously stirred
3
3
2
aqueous slurry of impregnated (unreduced) Pd/Al O as the
2
3
19
mixture of Pd and Bi solutions formed an insoluble precipitate.
Catalytic Reaction. A catalyst (100 mg) was placed in a 50
mL three-neck round-bottom flask equipped with a silicone
rubber septum, a reflux condenser, and a gas storage balloon (2
Pack k-point meshes of 2 × 3 × 1 [for Pd(111) and
PdCu(110)] and 2 × 2 × 1 [for PdZn(111)] were used. For
the calculation of methoxide and methylamide, surfaces were
modeled using (3 × 4) and (2 × 2) unit cells for Pd(111) and
PdZn(111), respectively, with a four-atom-thick layer and a 13
Å vacuum spacing between the two surfaces. The atomic
coordinates were fully relaxed, while the lattice constants were
fixed. Convergence criteria comprised (a) a self-consistent field
−1
L) and pretreated under a H stream (50 mL min ) at 400 °C
2
for 0.5 h using a mantle heater. After the pretreatment, dry Ar
−
1
(
20 mL·min ) was passed into the flask to replace the residual
2
H , and the flask was cooled to room temperature. A reaction
−6
mixture containing a solvent (p-xylene, 3 mL), amine (1.0
mmol), alcohol (1.0 mmol), and an internal standard (biphenyl,
tolerance of 2.0 × 10 eV/atom, (b) an energy tolerance of 2.0
5
−
× 10 eV/atom, (c) a maximal force tolerance of 0.05 eV/Å,
3
−
0.1 mmol) was added to the flask through the septum. The
and (d) a maximal displacement tolerance of 2.0 × 10 Å for
structure optimization and energy calculation. The transition
state (TS) search was performed using the complete linear
synchronous transit (LST)/quadratic synchronous transit
catalytic reaction was initiated by immersing the reaction
apparatus into a preheated oil bath under 1 atm of dry Ar. The
temperature of the oil bath was controlled to maintain an actual
reaction temperature of 110 °C. The products were quantified
by flame ionization detection gas chromatography (Shimadzu
GC-2025 equipped with an InertCap Pure-WAX capillary
column, GL Sciences). The turnover frequency was determined
as the consumption rate of amine (millimoles per hour) per
amount of exposed Pd (mmol) estimated by CO chem-
isorption. The initial consumption rates were typically obtained
below 30% conversion (conv.). Deuterated substrates (benzyl-
20,21
(QST) method.
An LST maximization was performed,
followed by an energy minimization in the directions
conjugating to the reaction pathway. The approximated TS
structure was subjected to a QST maximization with conjugate
gradient minimization refinements. This cycle was repeated
until a stationary point was found. The convergence criterion
for TS calculations was set to a root-mean-square force
tolerance of 0.10 eV/Å.
5
947
DOI: 10.1021/acscatal.6b01677
ACS Catal. 2016, 6, 5946−5953

ACS Catalysis
Research Article
The adsorption energy is defined using eq 1:
ΔEad = EA−S − (E + E )
Therefore, the prepared PdZn/Al O catalyst seems to have a
2
3
slightly Pd-rich PdZn phase (Pd0.56Zn0.44). The small deviation
of the lattice fringe spacing from the value in the literature
observed in Figure 2d may be attributed to the slightly enriched
Pd: the larger atomic radius of Pd (140 pm) compared to that
(1)
S
A
where EA−S is the energy of the slab with the adsorbate, E is
A
the total energy of the free adsorbate, and E is the total energy
S
24
of Zn (135 pm) increases the lattice constant.
of the bare slab.
We also performed Fourier transform infrared (FT-IR)
experiments with CO adsorption to characterize the surface of
the prepared catalysts (Figure 3). Pd/Al O presents some
RESULTS AND DISCUSSION
The Pd-based intermetallic compounds were prepared using an
■
2
3
peaks, which are assigned to the stretching vibration of CO
impregnation method with Al O as a support (Pd M /Al O ,
2
3
x
y
2
3
−1
−1
adsorbed in linear (2090 cm ) and bridging [1995 cm on
where M = Bi, Cu, Fe, Ga, In, Sn, Pd, or Zn). The crystallite
phases of the prepared catalysts were analyzed by XRD (Figure
). For most catalysts, the desired intermetallic phases were
formed with high phase purities. One-to-one alloy phases were
observed for the Pd−Cu and Pd−Fe catalysts.
Pd(100), 1950 cm− on Pd(111)] features. The contribution
of the bridging CO was dominant compared with that of the
linear-type CO on Pd/Al O . Conversely, for PdZn/Al O ,
1
25
1
2
3
2
3
linear-type CO was predominantly observed, indicating that the
Pd−Pd ensembles almost disappeared by the formation of the
intermetallic phase. The peak position of linear CO on PdZn/
26
Al O agreed with the values in the literature for PdZn and
2
3
was red-shifted compared with that on Pd/Al O . The latter
2
3
indicates that the Pd atoms in PdZn were slightly electron-
enriched by the charge transfer from Zn.
The prepared Pd-based catalysts were tested as catalysts for
alcohol-based N-alkylation of amines. First, aniline and benzyl
alcohol were used as model substrates. The obtained
conversions of benzyl alcohol and selectivities for the product
N-benzylaniline are listed in Table 1. For all catalytic reactions,
the products were typically N-benzylaniline, N-benzylideneani-
line (i.e., the intermediate imine), and toluene. Only a
negligible amount of benzaldehyde (<0.5%) was detected for
each reaction period, indicating that the condensation of
benzaldehyde and aniline to N-benzylideneaniline was
sufficiently fast. Monometallic Pd/Al O exhibited moderate
2
3
aniline conversion and selectivity (sel.) for N-benzylaniline
entry 1). A longer reaction time resulted in a slight increase in
(
Figure 1. X-ray diffraction (XRD) patterns of Pd-based bimetallic
catalysts supported on Al O . For the sake of clarity, diffraction of the
aniline conversion and significant reduction in product
selectivity (entry 10). The limited aniline conversion can be
attributed to the hydrogenolysis of benzyl alcohol to toluene
occurring as a side reaction (Figure S1a), which irreversibly
wasted the alkylating agent and hydrogen. The decrease in
product selectivity may be explained by the fragmentation of N-
benzylaniline into toluene and aniline. This was demonstrated
by a control experiment using N-benzylaniline alone as a
substrate, which resulted in the formation of toluene and
aniline in a 1:1 molar ratio (Figure S1b). Here, N-benzylaniline
itself acted as a hydrogen source, and dehydrogenated N-
benzylideneaniline was formed. Thus, undesired C−O and/or
C−N bond cleavage significantly occurred when monometallic
2
3
Al O support was subtracted from the original patterns. Crystallite
2
3
sizes estimated using the Scherrer equation are also shown.
Further characterization was performed for the catalysts that
presented broad diffraction peaks. Panels a and b of Figure 2
show the TEM image of PdZn/Al O and the size distribution
2
3
of the nanoparticles, respectively. The particle sizes range from
to 7 nm, and the mean volume diameter is 6.0 nm, which are
3
consistent with the crystallite size estimated using XRD (5 nm).
Panels c and d of Figure 2 show a high-resolution TEM image
of a single nanoparticle and the magnification of the region
(
designated by a yellow square in Figure 2c), respectively.
These images clearly show lattice fringes [spacing of 2.22 nm
Figure 2d, yellow lines)] that correspond to those of the
PdZn(101) planes [2.19 nm, ICDD-PDF 00-006-0620 (Figure
d, light blue lines)]. Moreover, the observed atomic arrays
closely match the corresponding crystal structure viewed along
the [11 ] direction.
Pd/Al O was used. Although Pd Ga and PdCu displayed
2 3 2
selectivities and yields higher than those of monometallic Pd
(entries 3 and 4, respectively), the product yields did not
increase even under much longer reaction times (entries 11 and
12, respectively). This is also because of the formation of
toluene, which limited aniline conversion and N-benzylidenea-
niline hydrogenation (Figure S1c). Thus, for this reaction,
undesired C−O and/or C−N bond scissions significantly
influence the product selectivity and should be avoided to
achieve high yields. Intermetallic catalysts with Fe, Sn, In, Bi,
and Pb exhibited considerably low catalytic performances
(entries 5−9, respectively). In contrast, PdZn exhibited
excellent catalytic activity and selectivity (entry 2 and Figure
S1d) with minimal toluene formation during the reaction
period (Figure S1d), allowing almost stoichiometric conversion
to the product N-benzylaniline.
(
2
̅ ̅
1
Chemical analysis was also performed for the prepared
PdZn/Al O catalyst, which contains a component that has a
2
3
low boiling point (Zn; 907 °C), using ICP-AES. A small
portion of Zn was lost (1:0.79 Pd:Zn) during the catalyst
preparation, probably because of the high vapor pressure and
the high reduction temperature used (800 °C). The loss of Zn
2
2
from PdZn at high temperature has been previously reported.2
3
According to the phase diagram of the Pd−Zn system,
intermetallic PdZn has a composition width (Pd, 44−63%).
5
948
DOI: 10.1021/acscatal.6b01677
ACS Catal. 2016, 6, 5946−5953

ACS Catalysis
Research Article
Figure 2. (a) Transmission electron microscopy (TEM) image of PdZn/Al O and (b) size distribution of the nanoparticles. (c) High-resolution
2
3
TEM image of a single nanoparticle. (d) Close-up of the region marked by the yellow square in panel c. The crystal structure of PdZn viewed along
the [11 ] direction is overlapped with the TEM image. Yellow and light blue lines represent experimental and literature values for d spacings,
respectively. (e) Fast Fourier transform (FT) of the image in panel c. (f) Rhombitruncated octahedron model of a PdZn nanoparticle with {110} and
101} facets viewed along the [11 ] direction.
̅ ̅
1
{
̅ ̅
1
Table 1. Catalytic Performances of Pd-Based Intermetallic
Compounds Supported on Al O in N-Alkylation of Aniline
2
3
a
with Benzyl Alcohol
conv. (%)
entry
catalyst
time (h)
1
2
sel. of 3 (%) yield of 3 (%)
1
2
3
4
5
6
7
8
9
Pd
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
5
77
100
64
45
48
9
50
99
52
45
27
7
29
99
65
72
19
36
9
22
99
42
32
9
PdZn
Figure 3. FT-IR spectra of CO chemisorbed on the Pd/Al O and
2
3
Pd Ga
2
PdZn/Al O catalysts.
2
3
PdCu
PdFe
Next, the catalysts exhibiting moderate to high aniline
conversions were tested for N-alkylation of 4-methylbenzyl-
amine, a highly reactive amine, with benzyl alcohol (Table 2;
the time course of the product yields is shown in Figure S2). In
this reaction, the desired N-alkylation yields an asymmetric
amine (1) and the corresponding imine (2), whereas the
undesired amine dimerization forms a symmetric amine (3)
and the corresponding imine (4). The preference for N-
alkylation over amine dimerization can be quantified by the rate
ratio of these reactions, r /r , where r and r represent the N-
Pd Sn
13
3
2
0
PdIn
PdBi
7
3
1
3
3
8
<1
<1
7
Pd Pb
9
3
3
<1
7
1
3
10
11
12
13
Pd
100
100
99
100
58
75
82
66
Pd Ga
2
3
49
64
35
49
63
35
PdCu
PdFe
7
4
a
Reaction conditions: amine, 1.0 mmol; alcohol, 1.0 mmol; solvent, 3
mL (p-xylene); PdZn/Al O , 100 mg (2.8 mol % Pd); temperature,
A
D
A
D
2
3
alkylation rate (rate of formation of 1 and 2) and amine
dimerization rate (rate of formation of 3 and 4), respectively.
Note that such a preference cannot be evaluated using
benzylamine and benzyl alcohol because the N-alkylation
products are identical to the dimerization products. In the case
of monometallic Pd, N-alkylation and dimerization proceeded
at comparable rates, affording an r /r value close to unity
1
10 °C; atmosphere, 1 atm of Ar.
because the involvement of amine dimerization and other side
reactions such as toluene formation remains significant.
Conversely, PdZn exhibited excellent selectivity for the desired
amine 1 and an r /r value that is 1 order of magnitude higher
A
D
A
D
(Table 2, entry 1). This indicates that monometallic Pd shows
than those of Pd Ga and PdCu (Table 2, entry 4). Thus, the
2
little preference for N-alkylation. Other byproducts such as
toluene were also formed, as observed in the aniline N-
PdZn catalyst allows preferential activation of the alcohol in the
presence of a highly reactive amine.
alkylation. Although Pd Ga and PdCu showed relatively high
The substrate scope of the PdZn/Al O catalyst in the
2
2
3
r /r values (Table 2, entries 2 and 3, respectively), the
selectivities for the desired amine 1 were moderate. This is
alcohol-based N-alkylation of amines was investigated, as
summarized in Table 3. The reaction between benzyl alcohol
A
D
5
949
DOI: 10.1021/acscatal.6b01677
ACS Catal. 2016, 6, 5946−5953

ACS Catalysis
Research Article
Table 2. Catalytic Performances of Pd-Based Catalysts in N-
Alkylation of 4-Methylbenzylamine with Benzyl Alcohol
catalytic performance (Figure S3). Although a decrease in the
catalytic performance (0.5 h, 55% conv. and 90% sel.) was
apparent in the fifth run, increasing the reaction time led to a
high product yield (2 h, 92% conv. and 92% sel.). Inductively
coupled plasma atomic emission spectrometry analysis revealed
that a small amount of Zn (∼6%) was leached during the four
initial catalytic runs, which may be the reason for the loss of
catalytic performance. In addition, aliphatic (primary and
secondary) and heteroaromatic alcohols could also be used as
alkylating agents for anilines with excellent catalytic perform-
ances (entries 4−6). For aliphatic primary amines, alkylation by
benzyl and aliphatic (primary and secondary) alcohols was
selectively catalyzed with high yields (entries 7−9). Alkylation
of secondary amines to the corresponding tertiary amines
proceeded successfully (entry 10). Benzylic amines, which are
highly reactive amines, were successfully converted to the
corresponding N-alkylated amines with good to excellent yields
a
selectivity (mol %)
N
time
conv.
(%)
entry catalyst (min)
1
2
3
4
others r /r
A D
1
2
3
4
Pd
15
30
45
40
100
97
29
61
63
92
1
4
2
3
23
9
5
5
4
1
42
18
17
2
1.5
7.7
Pd Ga
2
PdCu
PdZn
94
8
11
104
100
2
a
Reaction conditions are identical to those cited in the footnote of
Table 1.
(
entries 11−14). Alkylation of benzylamine with an aliphatic
alcohol (i.e., a highly reactive amine and a less reactive alcohol),
which is an extremely challenging reaction, resulted in a
moderate yield when triple the amount of alcohol was used
Table 3. N-Alkylation of Various Amines Using PdZn/Al O
Catalysts
2
3
a
(
1
entry 15). Note that most reactions in Table 3 (except entry
5) proceeded with equimolar amounts of amine and alcohol.
Thus, an efficient catalytic system for N-alkylation of amines
with high catalytic performances, a wide substrate scope, and
high atom efficiency was developed using intermetallic PdZn/
Al O .
time
(h)
conv.
(%)
sel.
entry
1
R1
R2
R3
(%)
Bn
Bn
Bn
Ph
Ph
Ph
Ph
Ph
Ph
H
H
H
H
H
H
H
H
H
0.5
2.5
24
6
100
100
100
92
99
>99
97
b
2
3
2
c
Subsequently, a kinetic study of the PdZn intermetallic
compound (Table 4) was performed. For PdZn, positive kinetic
isotope effect (KIE) values were obtained upon deuteration at
the α-C−H bond of benzyl alcohol for N-benzylation of aniline
3
d
4
d
n-C H
>99
94
8
17
5
c-C H
8
90
6
11
6
7
furfuryl
Bn
4
91
85
(
k /k = 2.2) and at the α-C−H bond of benzylamine for its
H
D
n-C H
2
97
95
8
17
17
17
17
d
dimerization (k /k = 2.3). This indicates that activation of the
H D
8
d
n-C H
n-C H
7
98
96
8
17
8
α-C−H bond is the rate-determining step for each reaction.
The corresponding activation energies were estimated by
Arrhenius-type plots (Figure S4), affording values of 77 and 96
kJ mol− for benzyl alcohol and benzylamine, respectively.
These results demonstrate that alcohol is activated much more
easily than amine on the surface of PdZn. Although a similar
kinetic study was performed for monometallic Pd, no KIE was
observed upon deuteration at the α-C−H bond of benzylamine,
suggesting that the rate-determining step of amine dimerization
is not α-C−H bond activation. We previously reported that, for
oxidative dehydrogenation of amine on monometallic Pd,
desorption of the product imine is the rate-determining step
because of its strong adsorption property. Similarly, product
desorption can also be the rate-determining step of benzyl-
amine dimerization over Pd. Thus, Arrhenius-type analysis of
the Pd catalyst cannot compare the activation energies of α-C−
H bond scissions. Therefore, we performed a first-principle
calculation for the C−H bond activation steps over a Pd(111)
surface using methoxide and methylamide as model adsorbates.
Although these adsorbates on Pd(111) have been theoretically
9
c-C H
n-C H
2
96
94
6
11
8
d
1
1
1
1
1
1
0
1
2
3
4
5
n-C H
n-C H
n-C H
7
92
>99
95
8
17
8
8
17
Bn
Bn
Bn
Bn
Bn
H
H
H
H
H
1
100
100
100
100
100
1
4-F-Bn
4-Me-Bn
4-MeO-Bn
Bn
0.75
0.67
0.75
2
88
92
67
d,e
n-C H
48
8
17
a
Reaction conditions: amine, 1.0 mmol; alcohol, 1.0 mmol; solvent, 3
mL (p-xylene); PdZn/Al O , 100 mg (2.8 mol % Pd); temperature,
2
3
b
c
1
1
3
10 °C; atmosphere, 1 atm of Ar. Temperature, 75 °C. Substrates,
d e
0 mmol; solvent (p-xylene), 30 mL. PdZn/Al O , 200 mg. Alcohol,
2
3
.0 mmol.
and aniline proceeded stoichiometrically even at a lower
temperature (75 °C, entry 2). The gram-scale reaction was also
well-catalyzed (entry 3), affording a turnover number of 2380
based on the Pd dispersion measured by CO adsorption
(
7.1%).
Recycling experiments revealed that the PdZn/Al O catalyst
2
3
can be used at least four times without significant loss of
Table 4. Summary of the Results of a Kinetic Study of N-Alkylation and Amine Dimerization over PdZn/Al O and Pd/Al O
3
2
3
2
reaction order
kinetic isotope effect
catalyst
PdZn
substrate
[alcohol]
[amine]
Ea (kJ mol−1)
deuteration
k /k
rate-determining step
H
D
PhNH + BnOH
0.01
−
0.45
0.20
−
−0.25
−
77
96
56
78
α,α-d -BnOH
2.2
2.3
2.8
1.0
C−H bond activation (BnOH)
C−H bond activation (BnNH2)
C−H bond activation (BnOH)
imine desorption
2
2
BnNH2
α,α-d -BnNH2
2
Pd
PhNH + BnOH
α,α-d -BnOH
2
2
BnNH2
−
α,α-d -BnNH2
2
5
950
DOI: 10.1021/acscatal.6b01677
ACS Catal. 2016, 6, 5946−5953

ACS Catalysis
Research Article
studied and the stable adsorption geometries have also been
2
7−31
clarified,
there are no reports about the comparison of C−
H bond activation energies for methoxide and methylamide
under the same conditions. Figure 4 shows the energy diagram
Figure 5. Calculated adsorption energy of oxygen (E ) and nitrogen
O
(
EN) atoms on various Pd-based surfaces. Optimized structures of the
surfaces with oxygen atoms adsorbed on the hollow sites are shown.
Figure 4. Energy diagrams for C−H bond activation of methoxide and
methylamide over a Pd(111) surface. All energies are relative to gas-
phase methanol or methylamine with the bare slab. The total energy of
bare slab and gas-phase methanol or methylamine was set to zero. For
each model, a hydrogen atom was adsorbed on an fcc hollow site at
the opposite surface of the slab to balance the number of atoms.
the monometallic Pd surface is azophilic. This trend is also
observed for methoxide and methylamide adsorbed on
Pd(111), as shown in Figure 4. Energetically, the methylamide
−1
adsorbate is favored over methoxide adsorbate by 34 kJ mol .
Conversely, for PdCu(110), E was slightly higher than EN
O
−1
(
ΔE
= 22 kJ mol ). For PdZn(111), E was much higher
O−N O
−
1
of C−H bond activation from methoxide to formaldehyde and
from methylamide to methanimine (TS structures are shown in
Figure S5). As the most stable adsorption sites, fcc hollow,
than E (ΔE
= 86 kJ mol ), which is consistent with the
O−N
N
preferential adsorption of alcohol over amine, as mentioned
above (Table 4). Thus, the adsorption affinity was drastically
modified by the formation of intermetallic phases. This may be
attributed to the intrinsic oxophilic character of Zn or Cu,
which is contrary to the azophilic character of Pd. The oxophilic
character of Zn atoms in the PdZn phase has also been
2
bridge, and μ−η sites were selected for methoxide,
methylamide, and the dehydrogenated products (formaldehyde
and methanimine), respectively. The C−H bond activation
−
1
−1
energy for methylamide (65 kJ mol ) was 24 kJ mol lower
−1
28
than that for methoxide (89 kJ mol ), supporting the fact that
amine is more easily activated than alcohol over a monometallic
Pd surface. On the basis of these results, it can be said that the
formation of a PdZn intermetallic phase inverts the C−H bond
activation favorability to alcohol and amine.
reported in the literature of computational studies. As is
evident in Figure 6, the ΔEO−N values for Pd, PdCu, and PdZn
Besides the aforementioned C−H bond activation properties,
the relative adsorption affinity of the substrates was also
considered. The dependence of the reaction rate of aniline N-
benzylation on the substrate concentration was examined, as
listed in Table 4 (the corresponding plots are shown in Figure
S6). When monometallic Pd was used as a catalyst, positive and
negative reaction orders were observed for alcohol and amine,
respectively, suggesting competitive adsorption of amine with
alcohol. In contrast, for PdZn, both reaction orders were nearly
zero. Considering the rate-determining step being α-C−H
bond activation of benzyl alcohol, the saturation coverage of
alcohol on the catalyst surface is likely. These results indicate
that alcohol adsorption is much more favorable than amine
adsorption on the PdZn surface. Figure 5 summarizes the
Figure 6. Relation between r /r and ΔEO−N.
A D
display a strong positive correlation with their r /rD ratios
N
obtained in the N-alkylation. This indicates that the adsorption
favorability of alcohol over amine plays a significant role in
product selectivity. Thus, the obtained results demonstrate that
the activation of alcohol over PdZn is preferred over the
activation of amines from kinetic and thermodynamic
perspectives, allowing highly efficient alcohol-based N-alkyla-
tion of amines.
calculated adsorption energies of oxygen (E ) and nitrogen
O
(
EN) atoms on the surfaces of Pd, PdCu, and PdZn. As the
surfaces that are the most stable and likely to be exposed, the
Pd(111), PdCu(110), and PdZn(111) planes were considered.
Pd-rich hollow sites (Pd fcc, Pd Cu, and Pd Zn hollow sites)
2
2
were chosen as the adsorption sites. In the case of the Pd(111)
surface, nitrogen showed an adsorption energy slightly higher
Intermetallic PdZn is considered to be an effective catalyst
for methanol steam re-forming with minimal byproduction of
CO. The positive role of the PdZn phase has been extensively
−1
22
than that of oxygen (ΔE
= E − E = −23 kJ mol ). This
O−N
O
N
indicates the intrinsic adsorption favorability of nitrogen atoms
to monometallic Pd compared with that of oxygen atoms; i.e.,
studied and proposed as inhibiting the dehydrogenation of the
22
formaldehyde intermediate to CO, which differs from the key
5
951
DOI: 10.1021/acscatal.6b01677
ACS Catal. 2016, 6, 5946−5953

ACS Catalysis
Research Article
role suggested in this study. Besides this reaction, Pd−Zn
bimetallic catalysts, including Pd/ZnO, have been widely
examined and have shown to be effective for various catalytic
REFERENCES
■
(
1) Lawrence, S. A. Amines: synthesis, properties and applications;
Cambridge University Press: Cambridge, U.K., 2004.
2) He, L.; Lou, X. B.; Ni, J.; Liu, Y. M.; Cao, Y.; He, H. Y.; Fan, K.
32−37
reactions, such as selective hydrogenation,
water−gas shift
(
3
8−40
41−43
reaction,
partial oxidation of methanol,
hydro-
N. Chem. - Eur. J. 2010, 16, 13965−13969.
(3) Shimizu, K.; Nishimura, M.; Satsuma, A. ChemCatChem 2009, 1,
497−503.
44
45
formylation of ethylene, and esterification. However, no
reactions involving nitrogen have been reported, highlighting
the novelty of PdZn catalysis for N-alkylation of amines.
(4) Likhar, P. R.; Arundhathi, R.; Kantam, M. L.; Prathima, P. S. Eur.
J. Org. Chem. 2009, 2009, 5383−5389.
(
5) He, W.; Wang, L. D.; Sun, C. L.; Wu, K. K.; He, S. B.; Chen, J. P.;
CONCLUSION
■
Wu, P.; Yu, Z. K. Chem. - Eur. J. 2011, 17, 13308−13317.
(6) Corma, A.; Rodenas, T.; Sabater, M. J. Chem. - Eur. J. 2010, 16,
254−260.
In this study, the catalytic performance of Pd-based
intermetallic compounds in alcohol-based N-alkylation of
amines was investigated. PdZn/Al O was observed as a highly
(7) Shimizu, K.; Imaiida, N.; Kon, K.; Hakim Siddiki, S. M. A.;
Satsuma, A. ACS Catal. 2013, 3, 998−1005.
2
3
efficient heterogeneous catalyst with high catalytic performance,
a wide substrate scope, and high atom efficiency. The surface of
intermetallic PdZn is capable of selective activation of alcohols
over amines, which is necessary for alkylation of reactive amines
such as benzylamine. The unique catalytic performance of
PdZn originates from the fundamental change in the C−H
activation property and adsorption affinity of the substrates,
allowing preferential activation of alcohols over amines. The
incorporation of the oxophilic Zn into the azophilic Pd at an
atomic level and ordered fashion plays a significant role in this
drastic change, which has rarely been observed using conven-
tional monometallic or supported catalysts. To the best of our
knowledge, this is the first report of the rationalization of this
kind of selective activation at an atomic scale. Thus, the results
obtained provide not only an efficient catalytic system for N-
alkylation of amines but also an innovative concept for selective
molecular recognition and activation.
(
8) Peeters, A.; Claes, L.; Geukens, I.; Stassen, I.; De Vos, D. Appl.
Catal., A 2014, 469, 191−197.
(
(
(
(
3
(
9) Yan, T.; Feringa, B.; Barta, K. Nat. Commun. 2014, 5, 5602−5602.
10) Yan, T.; Feringa, B. L.; Barta, K. ACS Catal. 2016, 6, 381−388.
11) Ma, X.; Su, C.; Xu, Q. Top. Curr. Chem. 2016, 374, 1−74.
12) Furukawa, S.; Suga, A.; Komatsu, T. Chem. Commun. 2014, 50,
277−3280.
13) Furukawa, S.; Suga, A.; Komatsu, T. ACS Catal. 2015, 5, 1214−
1222.
(14) Furukawa, S.; Yoshida, Y.; Komatsu, T. ACS Catal. 2014, 4,
1441−1450.
(
15) Segall, M. D.; Lindan, P. J. D.; Probert, M. J.; Pickard, C. J.;
Hasnip, P. J.; Clark, S. J.; Payne, M. C. J. Phys.: Condens. Matter 2002,
4, 2717−2744.
16) Vanderbilt, D. Phys. Rev. B: Condens. Matter Mater. Phys. 1990,
1
(
4
(
1, 7892−7895.
17) Zhang, Y. K.; Yang, W. T. Phys. Rev. Lett. 1998, 80, 890−890.
(
18) Hammer, B.; Hansen, L. B.; Norskov, J. K. Phys. Rev. B: Condens.
Matter Mater. Phys. 1999, 59, 7413−7421.
ASSOCIATED CONTENT
(19) Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188−5192.
■
(20) Halgren, T. A.; Lipscomb, W. N. Chem. Phys. Lett. 1977, 49,
*
S
Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.6b01677.
2
25−232.
21) Govind, N.; Petersen, M.; Fitzgerald, G.; King-Smith, D.;
Andzelm, J. Comput. Mater. Sci. 2003, 28, 250−258.
22) Armbruster, M.; Behrens, M.; Fottinger, K.; Friedrich, M.;
Gaudry, E.; Matam, S. K.; Sharma, H. R. Catal. Rev.: Sci. Eng. 2013, 55,
89−367.
23) Massalski, T. B.; Okamoto, H.; Subramanian, P.; Kacprzak, L.;
(
(
Reaction profile, recycling test, kinetic studies, and
optimized structures (PDF)
2
(
Scott, W. W. Binary alloy phase diagrams; American Society for Metals
International: Materials Park, OH, 1986; Vol. 1.
AUTHOR INFORMATION
■
*
Corresponding Authors
(
24) Slater, J. C. J. Chem. Phys. 1964, 41, 3199−3204.
Department of Chemistry, School of Science, Tokyo Institute
of Technology, 2-12-1-E1-10 Ookayama, Meguro-ku, Tokyo
52-8551, Japan. E-mail: furukawa.s.af@m.titech.ac.jp. Tele-
phone: +81-3-5734-2602. Fax: +81-3-5734-2758.
Department of Chemistry, School of Science, Tokyo Institute
of Technology, 2-12-1-E1-10 Ookayama, Meguro-ku, Tokyo
52-8551, Japan. E-mail: komatsu.t.ad@m.titech.ac.jp. Tele-
(25) Rainer, D. R.; Wu, M. C.; Mahon, D. I.; Goodman, D. W. J. Vac.
Sci. Technol., A 1996, 14, 1184−1188.
1
(26) Lebarbier, V.; Dagle, R.; Conant, T.; Vohs, J. M.; Datye, A. K.;
Wang, Y. Catal. Lett. 2008, 122, 223−227.
(
27) Zhang, C. J.; Hu, P. J. Chem. Phys. 2001, 115, 7182−7186.
*
(28) Chen, Z. X.; Neyman, K. M.; Lim, K. H.; Rosch, N. Langmuir
2
(
004, 20, 8068−8077.
29) Jiang, R. B.; Guo, W. Y.; Li, M.; Fu, D. L.; Shan, H. H. J. Phys.
1
phone: +81-3-5734-3532. Fax: +81-3-5734-2758.
Chem. C 2009, 113, 4188−4197.
Present Address
S.F.: Institute for Catalysis, Hokkaido University, N-21, W-10,
Sapporo 001-0021, Japan.
(30) Gu, X. K.; Li, W. X. J. Phys. Chem. C 2010, 114, 21539−21547.
†
(31) Liu, J. H.; Lv, C. Q.; Guo, Y.; Wang, G. C. Appl. Surf. Sci. 2013,
2
(
71, 291−298.
32) Sarkany, A.; Zsoldos, Z.; Furlong, B.; Hightower, J. W.; Guczi, L.
Notes
J. Catal. 1993, 141, 566−582.
(33) Iwasa, N.; Yoshikawa, M.; Arai, M. Phys. Chem. Chem. Phys.
The authors declare no competing financial interest.
2
(
2
(
002, 4, 5414−5420.
34) Iwasa, N.; Takizawa, M.; Arai, M. Appl. Catal., A 2005, 283,
55−263.
35) Lenarda, M.; Casagrande, M.; Moretti, E.; Storaro, L.; Frattini,
ACKNOWLEDGMENTS
■
This work was supported by JSPS KAKENHI Grant 26820350.
We deeply appreciate the Center for Advanced Materials
Analysis of the Tokyo Institute of Technology for the aid of
TEM observation.
R.; Polizzi, S. Catal. Lett. 2007, 114, 79−84.
(36) Semagina, N.; Grasemann, M.; Xanthopoulos, N.; Renken, A.;
Kiwi-Minsker, L. J. Catal. 2007, 251, 213−222.
5
952
DOI: 10.1021/acscatal.6b01677
ACS Catal. 2016, 6, 5946−5953

ACS Catalysis
Research Article
(
2
(
37) Tew, M. W.; Emerich, H.; van Bokhoven, J. A. J. Phys. Chem. C
011, 115, 8457−8465.
38) Bollmann, L.; Ratts, J. L.; Joshi, A. M.; Williams, W. D.;
Pazmino, J.; Joshi, Y. V.; Miller, J. T.; Kropf, A. J.; Delgass, W. N.;
Ribeiro, F. H. J. Catal. 2008, 257, 43−54.
(
39) Lebarbier, V.; Dagle, R.; Datye, A.; Wang, Y. Appl. Catal., A
2
(
3
010, 379, 3−6.
40) Wei, H.; Gomez, C.; Meyer, R. J. Top. Catal. 2012, 55, 313−
21.
(
41) Cubeiro, M.; Fierro, J. J. Catal. 1998, 179, 150−162.
(
42) Cubeiro, M.; Fierro, J. Appl. Catal., A 1998, 168, 307−322.
(
̈
̊
43) Agrell, J.; Germani, G.; Jaras, S. G.; Boutonnet, M. Appl. Catal.,
A 2003, 242, 233−245.
44) Arana, J.; Homs, N.; Sales, J.; Fierro, J.; Ramirez de la Piscina, P.
Catal. Lett. 2001, 72, 183−189.
45) Sanchez, A. B.; Homs, N.; Fierro, J.; Ramírez de la Piscina, P.
Catal. Today 2005, 107-108, 431−435.
(
̃
(
́
5
953
DOI: 10.1021/acscatal.6b01677
ACS Catal. 2016, 6, 5946−5953