Effect of an Alumina Phase on the Reductive Amination of 2-Propanol to Monoisopropylamine over Ni/Al2O3

Article abstract of DOI:10.1007/s10562-016-1695-8

Various single-phase aluminum oxides were prepared through the thermal decomposition of bayerite, boehmite, and gibbsite. Ni/γ-Al₂O₃ and Ni/δ-Al₂O₃ catalysts exhibited higher monoisopropylamine (MIPA) selectivity than the η-, θ-, and κ-Al₂O₃ supported Ni catalysts for the reductive amination of 2-propanol (IPA) in the presence of hydrogen and ammonia. FT-IR spectra after pyridine adsorption showed that a high number of Lewis acid sites could be correlated with enhancement in MIPA selectivity. Ni/η-Al₂O₃ and Ni/γ-Al₂O₃ catalysts exhibited the highest catalytic activity arising from differences in the metallic surface area. Both catalyst activity and selectivity with regards to reductive amination were strongly affected by the nature of the support.

Full text of DOI:10.1007/s10562-016-1695-8

Catal Lett
DOI 10.1007/s10562-016-1695-8
Effect of an Alumina Phase on the Reductive Amination
of 2-Propanol to Monoisopropylamine Over Ni/Al2O3
Jun Hee Cho1 Sang Hee An1 Tae-Sun Chang2 Chae-Ho Shin1
•
•
•
Received: 4 September 2015 / Accepted: 18 January 2016
Ó Springer Science+Business Media New York 2016
Abstract Various single-phase aluminum oxides were
prepared through the thermal decomposition of bayerite,
boehmite, and gibbsite. Ni/c-Al2O3 and Ni/d-Al2O3 cata-
lysts exhibited higher monoisopropylamine (MIPA) selec-
tivity than the g-, h-, and j-Al2O3 supported Ni catalysts
for the reductive amination of 2-propanol (IPA) in the
presence of hydrogen and ammonia. FT-IR spectra after
pyridine adsorption showed that a high number of Lewis
acid sites could be correlated with enhancement in MIPA
selectivity. Ni/g-Al2O3 and Ni/c-Al2O3 catalysts exhibited
the highest catalytic activity arising from differences in the
metallic surface area. Both catalyst activity and selectivity
with regards to reductive amination were strongly affected
by the nature of the support.
Graphical Abstract
Keywords 2-Propanol Á Alumina phase Á Reductive
amination Á Nickel Á Monoisopropylamine
Electronic supplementary material The online version of this
article (doi:10.1007/s10562-016-1695-8) contains supplementary
material, which is available to authorized users.
1 Introduction
& Chae-Ho Shin
chshin@chungbuk.ac.kr
Alkyl amines are used widely as intermediates for the
synthesis of herbicides, insecticides, and drugs. Their other
applications include rubber accelerators, corrosion inhibi-
tors, and surface-active agents. They can be obtained by the
reductive amination of aliphatic alcohols or carbonyl
compounds using ammonia in the presence of hydrogen.
However, yields and selectivities are rather low in the
synthesis of amines from their corresponding alcohols. The
production of primary amines from alcohols and ammonia
would be a very desirable process if it could be performed
Jun Hee Cho
dnvk@nate.com
Sang Hee An
anll59@nate.com
Tae-Sun Chang
tschang@krict.re.kr
1
Department of Chemical Engineering, Chungbuk National
University, Cheongju, Chungbuk 28644, Korea
2
Environment & Resources Research Center, Korea Research,
Institute of Chemical Technology, Daejon 34102, Korea
123

J. H. Cho et al.
with high selectivity for the amines [1–5]. Ni/Al2O3 cata-
lysts have been used widely in CH4 reforming, glycerol
steam reforming, and CO methanation [6–9].
filtered and washed with deionized water. The obtained
precipitate was dried in an oven at 100 °C for 24 h.
Boehmite (Sasol) and gibbsite (Samchun Chem.) were used
as purchased. We prepared g-Al2O3 through the calcina-
tion of bayerite at 600 °C for 4 h. j-Al2O3 was obtained
after gibbsite was calcined at 1000 °C for 4 h. We prepared
d-Al2O3 through the calcination of boehmite at 950 °C.
h-Al2O3 was obtained after calcination of c-Al2O3 (Sasol)
at 1050 °C for 2 h.
Synthesis of monoisopropylamine (MIPA) via reductive
amination of iso-propanol (IPA) with ammonia is per-
formed over a supported metal catalyst, typically nickel or
cobalt on silica and alumina, at temperatures of
190–250 °C and pressures of 18–50 bar [9]. The conver-
sion of IPA into MIPA includes three major reaction steps:
(1) dehydrogenation to form acetone, (2) condensation with
ammonia to form an imine, and (3) hydrogenation to form
MIPA. Each intermediate and MIPA can undergo various
side reactions such as condensation, decarbonylation, dis-
proportionation, and hydrogenolysis [10].
2.2 Catalyst Preparation
Commercial c-Al2O3 (Procatalyse) was used as a support.
g-Al2O3, j-Al2O3 [18], d-Al2O3, and h-Al2O3 were pre-
pared as described above and utilized as supports. Ni/Al2O3
catalysts were prepared by incipient wetness impregnation
using Ni(NO)3Á6H2O (Sigma-Aldrich) solution. The con-
centration of the impregnation solution was adjusted to
obtain a nickel loading of 17 wt%. The impregnated cat-
alysts were dried at 100 °C overnight and subsequently
calcined at 500 °C for 2 h in a muffle furnace under
flowing air.
The reductive amination of alcohols over nickel-based
catalysts has been studied using supports such as alumina,
silica, silica-alumina, zirconia, and zeolite [11–13]. Among
the various supports, Al2O3 has reasonably good mechan-
ical properties due to its thermal stability [14, 15]. The
choice of alumina as the catalyst support material also
affects activity and selectivity in the reductive amination
[6]. According to earlier studies, these support effects may
be due to changes in metal particle sizes and morphologies,
or interactions between the carrier and the supported metal.
In addition, amine selectivity is influenced by the acid–
base properties of the support [16–20]. Shimizu et al. [19]
investigated the direct synthesis of primary amines from
2-octanol with ammonia over Ni based catalyst showing
that the support properties such as acid–base sites influ-
enced to selectivity of primary amine.
2.3 Catalyst Characterization
Supports and catalysts were degassed for 6 h at 250 °C
before measurement and their Brunauer–Emmett–Teller
(BET) surface areas, pore volumes, and pore size distri-
butions were determined by N2 physisorption at -196 °C
using Micromeritics ASAP2020 apparatus. Surface areas
were calculated in the range of P/Po = 0.03–0.2. Pore size
distributions were calculated from the desorption branch
using Barrett–Joyner–Halenda (BJH) formulism. X-ray
diffraction patterns were obtained on a Rigaku diffrac-
tometer using Cu Ka radiation operating at 40 kV and
40 mA, with a scanning rate of 5 degree min-1 from 20° to
80°, to identify the crystalline phases of supports and cat-
alysts. Calcined catalysts, as well as those reduced at
600 °C for 3 h under H2 flow (50 cm3 min-1) followed by
passivation with 1.0 % O2/He for 0.5 h at room tempera-
ture (RT), were characterized separately to identify the
crystalline phases of NiO and Ni metal. H2-temperature-
programmed reduction (H2-TPR) using a Micromeritics
AutoChem II 2920 instrument determined the reducibility
of nickel oxides. Before measurement, each sample (0.1 g)
was pretreated under flowing He (50 cm3 min-1) up to
400 °C and held at that temperature for 1 h, to remove
adsorbed water and other contaminants, before being
cooled to 30 °C under flowing He. Reducing gas contain-
ing 5 % H2/Ar mixture (50 cm3 min-1) was then passed
over while heating to 800 °C at a rate of 10 °C min-1. A
thermal conductivity detector (TCD) determined the
amount of H2 consumed.
In our previous work, we examined the effects of nickel
loading in the reductive amination of IPA over Ni/c-Al2O3
catalysts and proposed that 17 wt% Ni/c-Al2O3 catalysts
were the most active [21]. No comparison work has been
conducted for the reductive amination of IPA over Ni
catalysts supported on alumina with different crystalline
phases. The objective was to study the interaction of nickel
particles with different alumina phases and to explore any
relationships between catalyst performance and support
properties.
2 Experimental
2.1 Support Preparation
Various single-phase aluminum oxides were prepared
through thermal decomposition of bayerite, boehmite, and
gibbsite. Bayerite was prepared by the precipitation method
at room temperature. NH4OH (Samchun, C28 %) was
added slowly until 1.0 M Al(NO3)2Á6H2O (Junsei) solution
reached pH 10, which was then maintained for 10 days
with vigorous stirring. The resultant white precipitate was
123

Effect of an Alumina Phase on the Reductive Amination of 2-Propanol to Monoisopropylamine…
The dispersion of nickel and the nickel surface area
were measured by H2 chemisorption at 100 °C under static
conditions using a Micromeritics ASAP 2020C instrument
equipped with a high-vacuum pump, providing a vacuum
of 10-6 Torr. Prior to adsorption experiments, each 0.3 g
sample was reduced at 600 °C for 3 h. H2 chemisorption
uptakes were separately determined as the difference of
two successive isotherm measurements. Metal dispersion
and the nickel metal surface area were calculated assuming
a H/Ni stoichiometry of 1. The degree of reduction was
determined by O2 titration using the following equation:
[amount of O2 consumed]/[theoretical amount of H2 con-
sumed assuming nickel oxides fully reduced (mmol H2;
NiO ? H2 ? Ni ? 2H?)] 9 100), and used to correct the
nickel particles sizes [22, 23].
FIPAreacted
Conversionð%Þ ¼
 100%
F
IPAfed
FiCi
iComponent selectivityð%Þ ¼
 100 %
RFiCi
À
Á
À
Á
FIPA mol hÀ1 Â MWIPA g molÀ1
Quantity of catalystðgÞ
À
Á
WHSV hÀ1
¼
where FIPA and Fi are molar flow rates of IPA and i prod-
ucts which are MIPA, diisopropylamine (DIPA), acetone,
respectively. MWIPA is a molecular weight of IPA. Ci is a
carbon number of i products.
3 Results and Discussion
The surface acidity of the Ni/Al2O3 catalysts reduced at
600 °C was measured with a Biorad 175C FT-IR spec-
trometer after pyridine adsorption. A self-supported sample
was prepared by pressing, and then treated under flowing
He (100 cm3 min-1) and H2 (20 cm3 min-1) at 400 °C for
1 h in situ in an IR cell. After cooling to 50 °C, the pre-
treated samples were exposed to pyridine vapor, and then
evacuated for 0.5 h at 50 °C. After treatment, IR spectra
were obtained at ambient temperature.
3.1 Physical Properties of Alumina
X-ray diffraction patterns for various Al2O3 supports are
shown in Fig. 1. The alumina phases, i.e. h-Al2O3,
d-Al2O3, c-Al2O3, g-Al2O3 and j-Al2O3, were identified
The morphology of Ni/Al2O3 catalysts was examined
with an Ultra High Resolution FE-SEM (Carl Zeiss, Res-
olution : 0.8 nm at 15 kV) and the energy dispersive X-ray
Analysis (EDS) mapping of Ni/Al2O3 catalysts was carried
out with a JEOL JSM-7000F (Resolution; 133 eV).
2.4 Catalytic Activity Tests
All catalytic experiments were performed under atmo-
spheric pressure at continuous-flow in a fixed bed
microreactor. Prior to testing, the catalysts were activated
at 600 °C for 3 h under a flow of purified H2 (50 cm3
min-1) and kept at the desired temperature to establish a
standard operating procedure, allowing time for product
distribution to stabilize. Standard conditions involved the
following reaction parameters: T = 160 °C; weight hourly
space velocity (WHSV, h-1) = 4.3; and feed composition
(IPA/NH3/H2/N2; molar ratio) = 1/4/6/22.8. The partial
pressure of IPA was fixed at 3 kPa.
For the comparison of selectivities of MIPA over vari-
ous Ni/Al2O3 catalysts, space velocity was controlled in the
range 2.3–19.1 at 150 °C. To maintain constant partial
pressures of the reactants, nitrogen was used as a diluent.
The reaction products were analyzed on-line using a
Chrompack-CP-9001 gas chromatograph equipped with a
CP-Volamine capillary column (60 mm 9 0.32 mm) and a
flame ionization detector. The conversion of IPA, selec-
tivity and WHSV were defined as follows:
Fig. 1 X-ray diffraction patterns of alumina supports prepared by
thermal decomposition of bayerite, boehmite, and gibbsite
123

J. H. Cho et al.
based on the highest intensities. Each alumina support has
its own characteristic bulk crystal structure [16, 24, 25]. It
has been reported that c-Al2O3 and g-Al2O3 have similar
crystal structures and that g-Al2O3 has a sharper (111)
peak than c-Al2O3 [16], which is consistent with our
results. The summarized results of the surface area and
pore volume for Al2O3 supports are shown in Table 1, and
the BJH pore size distribution is shown in Fig. 2. The
surface area increased in the following order: j-Al2O3 \
h-Al2O3 \ d-Al2O3 \ g-Al2O3 \ c-Al2O3. All of the
supports exhibited hysteresis during the adsorption and
desorption processes (data not shown). The pore size dis-
tributions were narrow, with ranges limited to less than
30 nm, except for the j-Al2O3 support, where the pore size
distribution ranged from 20 to 100 nm. The g-Al2O3 and
c-Al2O3 supports exhibited bimodal pore types.
Fig. 2 Pore size distribution of various alumina supports used in this
study: h-Al2O3, d-Al2O3, g-Al2O3, c-Al2O3, and j-Al2O3
3.2 Characterization of Ni/Al2O3 Catalysts
18.0 nm were observed in the Ni/j-Al2O3 catalyst, and the
minimum size observed in the Ni/g-Al2O3 catalyst was
around 9.6 nm, with the exception of the Ni/c-Al2O3 sup-
port. This observation is consistent with the pore size
measurement from the BJH method. The smaller nickel
oxide particles deposited on the g-Al2O3 support was
attributed to the small pore sizes in the Ni/g-Al2O3 catalyst
[18, 26]. The Ni/c-Al2O3 catalyst, with a larger specific
surface, had a smaller average particle size. The reduced
Ni/Al2O3 catalysts showed the characteristics peaks for
metallic Ni (JCPDS 04-0850) and Al2O3. Before and after
the reduction, the alumina phase of all catalysts was nearly
the same. This indicated that the crystal phase of the alu-
mina supports had not changed during reduction at 600 °C
for 3 h. The diffraction peaks of Ni metal gradually
became sharper and more intense in the following order:
Ni/g-Al2O3 \ Ni/c-Al2O3 \ Ni/h-Al2O3 \ Ni/d-Al2O3 \
Ni/j-Al2O3. Among the catalysts examined, the Ni/g-
Al2O3 catalyst exhibited the smallest Ni particle size of
8.7 nm, whereas the Ni/j-Al2O3 catalyst had the largest Ni
particle size of 18.7 nm.
The physical properties of calcined Ni/Al2O3 are summa-
rized in Table 1. BET surface area, pore volume, and pore
size progressively decreased after impregnation and calci-
nation. The Ni/Al2O3 catalysts showed reduced pore vol-
ume and average pore size, which means there was
efficient deposition of nickel species in the inner pores of
the Al2O3 support. Although pore volume and size after
nickel impregnation decreased, the surface area of the Ni/
j-Al2O3 catalysts decreased slightly, from 23 to 22 m2 g-1
.
This was probably due to the deposition of large nickel
particles on the outer surface of the j-Al2O3 [26].
The XRD patterns of the various Ni/Al2O3 catalysts
calcined at 500 °C and reduced at 600 °C are shown in
Fig. 3. The calcined catalysts exhibited a characteristic
reflection peak at 2h = 43.2° due to the presence of the
NiO (JCPDS 47-1049) phase. The present XRD analysis
suggested that there was no new compound formed
between nickel oxide and the alumina. The average particle
size of NiO was estimated using the Scherrer equation. As
shown in Table 2, the average particle size of NiO ranged
from 9.5 to 18.0 nm. Larger NiO particle sizes of around
Table 1 Physical properties of alumina supports and Ni/Al2O3 catalysts
Catalysts
SBET (m2 g-1
)
Pore volume (cm2 g-1
)
Pore diameter (nm) Supports
SBET (m2 g-1
)
Pore volume (cm2 g-1
)
Pore diameter (nm)
Ni/h-Al2O3
Ni/d-Al2O3
83
93
0.353
0.256
0.253
0.472
0.159
17.1
11.1
9.6
109
0.519
0.362
0.369
0.699
0.205
19.0
12.3
10.3
14.4
35.8
117
143
194
23
Ni/g-Al2O3 105
Ni/c-Al2O3 144
13.1
28.0
Ni/j-Al2O3
22
123

Effect of an Alumina Phase on the Reductive Amination of 2-Propanol to Monoisopropylamine…
catalysts, it has been reported that nickel aluminate may
form, which hinders complete reduction of the nickel spe-
cies. Thus, in the present case, the third-peak can be asso-
ciated with a nickel oxide layer that contacts the alumina
surface [29, 30]. Reduction of the peak for the Ni/j-Al2O3
catalyst at low temperatures was probably due to larger NiO
particles and lower dispersion [14, 17]. The intensity and
area of the first reduction peak shifted at lower temperatures
in the following order: Ni/j-Al2O3 \ Ni/g-Al2O3 \ Ni/
c-Al2O3 \ Ni/h-Al2O3 \ Ni/d-Al2O3. The nickel-support
interaction depended on the type of support and nickel par-
ticle size. In addition, it was known that the size of nickel
oxide particles affects its reducibility [31]. Generally, as the
particle size of the active metal in the M/Al2O3 catalyst
(M = active metal, such as Co, Ni, and Fe) increased, the
intensity and area of the first reduction peak also tended to
increase and shift to lower temperatures [32]. However,
according to the XRD results presented in Table 2, the dif-
ference in NiO particle size for various Ni/Al2O3 catalysts
was not significant. The TPR profiles showed that different
alumina phases affected the reduction process. However, the
reduction of NiOx to Ni takes place at lower temperatures for
Ni/c-Al2O3, Ni/h-Al2O3, and Ni/d-Al2O3 catalysts com-
pared with the Ni/g-Al2O3 catalyst.
H2 chemisorption on Ni/Al2O3 catalysts provides valu-
able and quantitative information such as the number of
nickel metal sites, nickel dispersion, and its average par-
ticle size. The results of H2 chemisorption, including the
metallic surface area, metal dispersion, and particle size,
corrected using the reduction degree measured by O2
titration, are presented in Table 2. The nickel particles,
when corrected by considering the degree of reduction,
were smallest (ca. 8.7 nm) in Ni/g-Al2O3. The particle size
increased from 12.5 to 41.9 nm in the following order: Ni/
g-Al2O3 \ Ni/c-Al2O3 \ Ni/h-Al2O3 \ Ni/d-Al2O3 \ Ni/
j-Al2O3. Nickel particles sizes were in good agreement
with the results of the XRD analysis, which showed the
smallest Ni particles to be 8.7 nm, found in Ni/g-Al2O3,
and the largest to be 41.9 nm, found in Ni/j-Al2O3. By
comparing the O2 titrations of nickel oxide on different
alumina phases for Ni/Al2O3 catalysts, the reduction
degree was observed to be almost constant. The highly
reduced nickel metal surface area could be correlated with
enhancement of IPA conversion [4, 18]. Unreduced nickel
oxides are not active in the reductive amination of IPA
because they cannot catalyze the dehydrogenation-hydro-
genation process [11]. The reduced metal surface area of
nickel was largest for the Ni/g-Al2O3 catalyst at ca.
Fig. 3 XRD patterns of Ni/Al2O3 catalysts with various alumina
supports a calcined at 500 °C for 2 h, and b reduced at 600 °C for 3 h
The H2-TPR profiles of calcined Ni/Al2O3 catalysts with
different alumina phase supports are shown in Fig. 4. The
TPR data revealed that different alumina phase supports
influence the reduction behavior of the catalysts. The
reducibility of the catalysts was compared as part of the
evaluation of interactions between nickel and the supports.
The TPR profiles of the Ni/Al2O3 catalysts calcined at
500 °C revealed that three main reduction peaks were
observed with a temperature range of 300–450, 450–650,
and 650–750 °C with broad reduction profiles, except for the
Ni/j-Al2O3 catalyst. The first peak, around 430 °C, is gen-
erally attributed to the reduction of bulk NiO or NiO weakly
interacting with the support. The broad high temperature
reduction peak, at around 600 °C, can be ascribed to the
reduction of NiOx species that have stronger binding to the
support. A third peak, at around 760 °C, is thought to be from
the reduction of nickel aluminate based on literature prece-
dence [27, 28]. In the case of alumina-supported nickel
À1
5.0 m2 g
, which correlated with IPA conversion. Ni
catalyst
particle size in Ni/g-Al2O3 (8.7 nm) was shown to be
smaller, therefore providing a higher surface area and
123

J. H. Cho et al.
Table 2 Nickel particle size, surface area measured by H2 chemisorption; reduction degree measured by O2 titration
Catalysts XRD
O2
titration
H2 chemisorption
H2 TPR tempearature (°C)
Particle
size of
Particle
size of
Reduction Uncorrected
degree
Corrected
Metallic
surface
1 st
peak
(%)
2nd
peak
(%)
3rd peak
(%)
NiO (nm) Ni (nm) (%)a
area (m2
Dispersion Particle Dispersion Particle
(%)b
-1
(%)
size
(nm)
size
(nm)c
g
)
cat
Ni/h-
Al2O3
Ni/d-
Al2O3
Ni/g-
Al2O3
Ni/c-
Al2O3
Ni/j-
Al2O3
14.5
14.2
9.6
13.6
15.0
8.7
49.1
48.4
48.5
47.8
48.4
2.8
2.4
4.4
2.9
1.2
36.7
42.9
23.0
34.5
86.6
5.6
4.9
9.1
6.1
2.4
18.0
20.8
11.2
16.5
41.9
3.0
2.7
5.0
3.4
1.3
418 (44) 523 (47) 643 (9)
420 (41) 525 (46) 660 (12)
369 (20) 544 (77) 746 (3)
373 (13) 515 (75) 698 (12)
368 (89) 480 (11)
9.5
12.5
18.7
18.0
a
b
c
Calculated from O2 uptake
Corrected dispersion (D) = Surface Ni0 atom/total reduced Ni0 atom 9 100 = Surface Ni0 atom/(total Ni atom 9 reduced fraction) 9 100
Corrected Ni diamete = Uncorrected Ni diameter 9 reduced Ni fraction
Fig. 5 FT-IR spectra of Ni/Al2O3 catalysts with different alumina
phase supports after pyridine adsorption
Fig. 4 H2-TPR profiles of calcined Ni/Al2O3 catalysts with different
alumina phase supports
tion of coupled Lewis and Brønsted acid sites. A band
corresponding to a-pyridone species, representing the
presence of basic surface OH groups and acid–base pair
sites, also appeared at 1577 cm-1 [33–35]. The absence of
bands around 1540 and 1640 cm-1 indicates that no
Brønsted sites were present on the catalyst surface. The
total amount of acid sites increased in the following order:
Ni/j-Al2O3 \ Ni/g-Al2O3 \ Ni/h-Al2O3 \ Ni/d-Al2O3 &
Ni/c-Al2O3. We speculate that acidity had an important
role in the product distribution over various Ni/Al2O3
larger Ni dispersion in comparison to other Ni/Al2O3
catalysts.
The FT-IR spectra recorded after pyridine adsorption at
50 °C were used to obtain information on the nature and
amount of acid sites over various Ni/Al2O3 catalysts, as
shown in Fig. 5. The bands from pyridine protonated by
Brønsted acid sites appeared at 1540 cm-1, while bands
from pyridine coordinated to the Lewis acid sites, assigned
to coordinatively unsaturated Al3?, appeared at 1446 and
1605 cm-1. The band at 1488 cm-1 indicates the forma-
123

Effect of an Alumina Phase on the Reductive Amination of 2-Propanol to Monoisopropylamine…
catalysts, with high selectivity towards MIPA being asso-
ciated with surface acidity.
observed variation in conversion could be correlated with
the differences in reduced metallic surface area in the
catalysts. This was likely due to variations in reducibility
and nickel metal dispersion that accompanied higher
metallic surface areas [2], as shown by the H2-
chemisorption results (Table 2). The sums of the selectiv-
ities of MIPA, acetone, and DIPA were almost constant at
ca. 100 % under the tested operating conditions. Acetone
can be formed by dehydrogenation of IPA on metallic
surface. Small quantities of DIPA were formed by a series
reaction of MIPA with IPA by condensation/dehydration.
An additional reversible reaction is DIPA ? NH3 ? 2-
MIPA [37]. DIPA can therefore be recycled in the reduc-
tive amination process. The MIPA selectivity of the Ni/
Al2O3 catalysts was higher than acetone selectivity, except
for Ni/j-Al2O3. Among the catalysts, MIPA selectivity was
lowest for the Ni/j-Al2O3 catalyst. This was considered to
be due to the overspread of large nickel particles on the
outer surface of the j-Al2O3 support. This meant that no
acidic or basic sites were present on the j-Al2O3 support.
In the reductive amination mechanism for IPA, the first and
last redox processes are catalyzed by the metal. In addition,
the reaction of the acetone intermediate with NH3 to form
an imine does not require a metal catalyst, but can be
accelerated over an acid or base catalyst [11, 38].
Figure S1 shows the SEM images of Ni/Al2O3 catalysts.
SEM was examined to observe the morphology and parti-
cle size of NiO over various Al2O3 supports. NiO particles
of all catalysts showed a similar morphology as spherical
shape. The average particle size is 8.6, 12.3, 18.3, 19.6 and
40.9 nm for Ni/g-Al2O3, Ni/c-Al2O3, Ni/h-Al2O3, Ni/
d-Al2O3, and Ni/j-Al2O3 catalyst, respectively. The parti-
cle size increased from 8.6 to 40.9 nm in the following
order: Ni/g-Al2O3 \ Ni/c-Al2O3 \ Ni/h-Al2O3 \ Ni/d-
Al2O3 \ Ni/j-Al2O3. The dispersion of NiO is hardly
observed by SEM images because it is difficult to classify
between Ni and alumina. The energy dispersive X-ray
Analysis (EDS) mapping of Ni/Al2O3 catalysts with vari-
ous alumina supports was shown in Fig. S2. SEM/EDS
mapping clearly reveals the higher dispersion of metallic
Ni in Ni/g-Al2O3 and Ni/c-Al2O3 more than Ni/j-Al2O3.
The dispersion was in good agreement with the results of
the O2 titration.
3.3 Reductive Amination of IPA
Table 3 shows the effect of the alumina phase on the
reductive amination of IPA over Ni/Al2O3 catalysts.
Results obtained with 30 min on stream were considered
intrinsic activities. The catalysts showed no deactivation
during this time. In our previous study, the deactivation of
Ni/c-Al2O3 catalyst was investigated in the presence or
absence of hydrogen for the reductive amination of IPA
[21]. The intrinsic catalytic activity for reductive amination
of IPA under the reaction conditions studied here increased
in the following order: Ni/d-Al2O3 \ Ni/j-Al2O3 \ Ni/h-
Al2O3 \ Ni/c-Al2O3 \ Ni/g-Al2O3. Sewell et al. [36] and
Chary et al. [13] suggested that alcohol conversion in
reductive amination is proportional to the metal surface
area, which is directly associated with metal dispersion.
Although conversion was not directly proportional to the
metallic surface area in the Ni/Al2O3 catalysts, the
For the comparison study of MIPA selectivity over Ni/
Al2O3 catalysts, the space velocity was maintained in the
range 2.3–19.1 at 150 °C under identical reaction condi-
tions. MIPA selectivity was dependent on the acidity of
catalyst [6, 13], and the selectivity data was compared at
constant IPA conversion (around 30 %) after 60 min of
operation. The MIPA selectivities obtained are summarized
in Table 4. The MIPA selectivity increased in the follow-
ing order: Ni/j-Al2O3 \ Ni/g-Al2O3 \ Ni/h-Al2O3 \ Ni/
d-Al2O3 & Ni/c-Al2O3. This result was in good agreement
with the total number of acidic sites in the FT-IR spectra
after pyridine adsorption (Fig. 6). IPA is dehydrogenated to
acetone, subsequent conversion to carbinolamine by
nucleophilic addition with ammonia, which in its turn is
Table 3 Effect of alumina
Catalysts
Conversiona (%)
Selectivity (%)a
MIPA yield (%)
phases in the reductive
amination of IPA over Ni/Al2O3
catalysts
MIPA
Acetone
DIPA
Ni/h-Al2O3
Ni/d-Al2O3
Ni/g-Al2O3
Ni/c-Al2O3
Ni/j-Al2O3
35.9
27.9
77.6
62.1
34.9
73.1
75.3
77.2
78.2
6.7
20.3
19.2
15.0
15.0
91.6
6.6
5.5
7.8
6.8
1.7
26.2
21.0
59.9
48.6
2.3
Reaction conditions: T = 160 °C, WHSV = 4.3 h-1; feed compositions of IPA/NH3/H2/N2 (mol.
ratio) = 1/4/6/22.8
a
Both conversion and selectivities to MIPA, acetone and DIPA obtained at 30 min on stream
123

J. H. Cho et al.
Table 4 Product distributions
at similar conversions of IPA
over various Ni/Al2O3 catalysts
Catalysts
WHSV (h-1
)
Conversion (%)
Selectivity (%)
MIPA
Acetone
DIPA
Ni/h-Al2O3
Ni/d-Al2O3
Ni/g-Al2O3
Ni/c-Al2O3
Ni/j-Al2O3
4.3
2.3
29.3
30.1
31.0
32.6
28.0
74.3
81.7
54.0
82.0
5.6
21.3
13.7
46.0
16.9
90.5
4.3
4.6
–
19.1
11.1
4.8
1.1
2.2
Reaction conditions: T = 150 °C, feed compositions of IPA/NH3/H2/N2 (mol. ratio) = 1/4/6/22.8
optimal reaction conditions for Ni/c-Al2O3 catalysts [21].
The long-term stability of Ni/g-Al2O3 and Ni/c-Al2O3
catalysts was examined at 160 °C, with a space velocity of
2.3 h-1 and an R = [IPA]/[NH3]/[H2]/[N2] molar ratio of
1:8:6:18.8, in the reductive amination of IPA over Ni/
g-Al2O3 and Ni/c-Al2O3 catalyst. The total flow rate was
50 cm3 min-1. The long-term stability of Ni/g-Al2O3 and
Ni/c-Al2O3 catalysts during the reductive amination of IPA
is shown in Fig. 7. These catalysts, which showed the best
catalytic activity and highest MIPA selectivity among the
Fig. 6 Correlation between the Lewis acid band intensity and the
MIPA selectivity for the reductive amination of IPA over Ni catalysts
supported on various alumina phases
readily dehydrated to an imine [4]. Dehydration to the
imine is considered a good measure of acidic or acid–base
properties at the catalyst surface. We confirmed that no
Brønsted sites were present on the catalyst surface by FT-
IR spectra after pyridine adsorption. Kirumakki et al. [10]
suggested that Brønsted acid (H?) was found to be inactive
in the reductive amination of alcohol. This means that the
reaction proceeded at Lewis acid sites according to an E2
elimination mechanism. After adsorption of carbinolamine
at the Lewis acid site, the base can attack it before disso-
ciation, which could produce an iminium ion [39, 40]. It is
important to evaluate the strength of Lewis acid sites based
on the rate of dehydration to imine, which depends on the
amount of formed imine for high selectivity toward MIPA.
We speculated that high MIPA selectivity was due to the
increase in the amount of adsorbed carbinolamine mole-
cules, affected by the increase in the Lewis acid sites on the
catalyst surface.
In our previous work, we examined the effect of nickel
loading and reaction parameters (the effect of H2, NH3,
reaction temperature, and space velocity) on the reductive
amination of IPA over Ni/c-Al2O3 catalysts and proposed
Fig. 7 Long-term stability of a Ni/g-Al2O3 and b c-Al2O3 in the
reductive amination of IPA. Reaction conditions: T = 160 °C,
WHSV = 2.3 h-1
(mol%) = 1/8/6/18.8
;
feed
composition
of
IPA/NH3/H2/N2
123

Effect of an Alumina Phase on the Reductive Amination of 2-Propanol to Monoisopropylamine…
10. Kirumakki SR, Papadaki M, Chary KVR, Nagarajua N (2010) J
Mol Catal A 321:15
11. Zamlynny V, Kubelkova L, Baburek E, Jiratova K, Novakova J
(1998) Appl Catal A 169:119
12. Vedage GA, Emig LA, Li HX, Armor JN (1999) US Patent
5,917,092
catalysts studied here, exhibited constant conversion to,
and selectivity for, MIPA for up to 100 h on stream. These
catalysts showed good catalytic activity and stability, and
their excellent performance make the process economically
viable.
13. Chary KVR, Seel KK, Naresh D, Ramakanth P (2008) Catal
Commun 9:75
14. Rane S, Borg Ø, Yang J, Ryttera E, Holmen A (2010) Appl Catal
A 388:160
4 Conclusion
15. Favaro L, Boumaza A, Roy P, Ledion J, Sattonnay G, Brubach
JB, Huntz AM, Tetot R (2010) J Solid State Chem 183:901
16. Kim YH, Park ED (2010) Appl Catal B 96:41
17. Wang R, Li Y, Shi R, Yang M (2011) J Mol Catal A 344:122
18. Park JE, Park ED (2014) Korean J Chem Eng 31:1985
19. Shimizu K, Kenichi K, Onodera W, Yamazaki H, Kondo JN
(2013) ACS Catal. 3:112
20. Park JE, Kim BB, Park ED (2015) Korea J Chem Eng 32:2212
21. Cho JH, Park JH, Chang TS, Kim J-E, Shin C-H (2013) Catal Lett
143:1319
22. Wang L, Li D, Koike M, Koso S, Nakagawa Y, Xu Y, Tomishige
K (2011) Appl Catal A 392:248
23. Bae JW, Kim SM, Kang SH, Chary KVR, Lee YJ, Kim HJ, Jun
KW (2009) J Mol Catal A 311:7
24. Kamal M, Khalil S (1998) J Catal 178:198
25. Elaloui E, Pierre AC, Pajonk GM (1997) J Catal 166:340
26. Oh JH, Bae JW, Park SJ, Khanna PK, Jun KW (2009) Catal Lett
130:403
27. Yang R, Li X, Wu J, Zhang X, Zhang Z, Cheng Y, Guo J (2009)
Appl Catal A 368:105
Various alumina supports, such as h-, d-, g-, c- and
j-Al2O3, influence the physicochemical and catalytic
properties of nickel-based catalysts in the reductive ami-
nation of IPA. Ni/g-Al2O3 had the smallest nickel particle
size and demonstrated the best catalytic performance, due
to a high nickel metal surface area and high dispersion of
nickel particles. The highest number of Lewis acid sites on
the catalyst surface could be correlated with enhancement
in the MIPA selectivity. The Ni/g- and c-Al2O3 catalysts
showed good catalytic performance in the reductive ami-
nation of IPA to MIPA.
Acknowledgments This work was supported by the Energy Effi-
ciency and Resources of the Korea Institute of Energy Technology
Evaluation and Planning (KETEP) granted financial resource from the
Ministry of Trade, Industry and Energy, Republic of Korea (No.
2014021194).
28. Li C, Chen YW (1995) Thermochim Acta 256:457
29. Chen J, Ma Q, Rufford TE, Li Y, Zhu Z (2009) Appl Catal A
362:1
30. Ji Y, Zhao Z, Duan A, Jiang G, Liu J (2009) J Phys Chem C
113:7186
´ ´
31. Prieto G, Martınez A, Concepcion P, Tost RM (2009) J Catal
References
266:129
32. Park JY, Lee YJ, Khanna PK, Jun KW, Bae JW, Kim YH (2010)
J Mol Catal A 323:84
33. Milewska JF, Decyk P, Ziolek M (2011) Appl Catal A 393:215
34. Kim YT, Jung KD, Park ED (2010) Microporous Mesoporous
Mater 131:28
35. Kim YT, Jung KD, Park ED (2011) Appl Catal B 107:177
36. Sewell G, O’Connor C, Steen E (1995) Appl Catal A 125:99
37. Luyben WL (2009) Ind Eng Chem Res 48:10551
38. Fischer A, Maciejewski M, Burgi T, Mallat T, Baiker A (1999) J
Catal 183:373
39. Turek W, Krowiak A (2012) Appl Catal A 417–418:102
40. Aramendia MA, Borau V, Jimenez C, Marinas JM, Porras A,
Urbano FJ (1996) J Catal 161:829
1. Hayes KS (2001) Appl Catal A 221:187
2. Rausch AK, Steen E, Roessner F (2008) J Catal 253:111
3. Jeon HY, Shin C-H, Jung HJ, Hong SB (2006) Appl Catal A
305:70
4. Cho JH, Park JH, Chang TS, Seo G, Shin C-H (2012) Appl Catal
A 417–418:313
5. Gardner DA, Clark RT (1981) US Patent 4,255,357
6. Huang Z-Y, Xu C-H, Liu C-Q, Xiao H-W, Chen J, Zhang Y-X,
Lei Y-C (2013) Korean J Chem Eng 30:587
7. Seong M, Shin M, Cho J-H, Lee Y-C, Park Y-K, Jeon J-K (2014)
Korean J Chem Eng 31:412
8. Majidian N, Habibi N, Rezaei M (2014) Korean J Chem Eng
31:1162
9. Dobson ID, Lidy WA, Williams PS (1990) US Patent 4,912,260
123